Introduction. Global climate warming is exacerbating the problem of landscape fires due to increased evaporation of moisture from combustible materials, more frequent dry thunderstorms, extended fire season, and shifting boundaries of landscape zones. These processes pose a particular threat to the forest regions of Russia, especially the Republic of Sakha (Yakutia). Domestic and international researchers have found that year-to-year variations in average monthly surface air temperature (MAT) have a significant impact on variations in fire risks. The authors have previously proved the existence of a positive reverse causality between the indicators of forest fire frequency and temperature anomalies of the following year in many regions of Siberia. However, the significance of this connection at the level of individual uluses of Yakutia, the territories of which coincide with the areas of responsibility of fire departments, has not previously been assessed. This creates a gap in scientific knowledge. The aim of this research is to fill this gap by assessing the significance of this relationship for all uluses of the republic and checking its resistance to time shifts of the analyzed series.

Materials and Methods. The research was conducted using data from 2000–2024. ERA5 reanalysis was used as a source of information on MAT distribution at a height of 2 m above the studied territories, corresponding to the nodes of the 0.25° grid. Information from the Remote Monitoring Information System of the Federal Forestry Agency was used as factual material on the number of landscape fires in the territory of each ulus (district) of Yakutia and the total area of its sites covered by fire each year during the specified period. For each of the 34 uluses (districts) of Yakutia, we calculated the average MAT for their entire territory for the months from May to July, taking into account MAT information corresponding to points for which such information was provided in ERA5. The statistical relationships between interannual changes in the average MAT for the period under review and variations in forest fires indicators throughout Yakutia (one year ahead) were studied using correlation analysis for various time periods lasting 10–20 years. Linear trends were removed from the time series before performing the analysis. The significance of correlation coefficients was assessed using the Student's criterion with a confidence level of 95% or higher. The stability of these relationships was verified by examining their consistency when the series were shifted by one year and when the analyzed segments were varied between 10 and 20 years in length.

Results. The uluses (districts) of Yakutia were identified, where statistical relationships between interannual changes in the average MAT for May–July, with one-year-ahead variations in the number of landscape fires and the area of fire-affected areas of Yakutia from 2000 to 2023, were found to be significant at a confidence level ≥95%. These include uluses (districts) located both in the northern and western parts of the republic's territory, as well as in its central part. The stability of the identified relationships was proved to time shifts of the analyzed periods by units of years into the past and future, as well as to changes in the duration of time series segments within 10–20 years. It was also established that during the period from 2001 to 2023, the relationships under consideration gradually strengthened: the number of uluses and districts with reliability of conclusions ≥95% increased more than twice, and the number of territories with reliability ≥99% rose from zero to 16. The relationship between changes in MAT for the identified uluses (districts) and variations in the number of fires in the territory of Yakutia were more reliable than the relationships with variations in the area of its parts affected by fire.

Discussion. The results confirmed the existence of uluses (districts) on the territory of Yakutia, for which the influence on the interannual changes in MAT exerted by variations in the indicators of forest fires throughout Yakutia, which were 1 year ahead of them in time, was significant. The novelty consisted in identifying all uluses (districts) for which the connections between these processes were significant and resistant to time shifts. The revealed stability of the discovered relationships indicated the fundamental nature of the dependence: contamination of snow by particles of fire aerosols deposited on it reduced the albedo of the underlying surface covered with it, which accelerated melting, increased the temperature and evaporation rate, increasing fire risk. During the period under study, these relationships strengthened, which indicated the influence of climate warming on the activation of the positive reverse causality under consideration. Therefore, with further warming of the climate of Yakutia, they would increase even more. Variations in the number of landscape fires throughout Yakutia had a stronger effect on MAT changes for the identified uluses (districts) than variations in the burned area. The results obtained made it possible to use the indicators of forest fires throughout Yakutia of the previous year as predictors for long-term MAT forecasts for its identified areas (uluses).

Conclusion. Uluses (districts) in Yakutia have been identified, for which the statistical relationship between changes in forest burning rates throughout the republic and annual MAT variations in May — July are significant. With a reliability of ≥95% for such uluses (districts) for the period from 2015 to 2024, 23 were identified, and with a reliability of ≥99%, 16 were identified. The stability of these relationships to time shifts and the duration of time series have been proved. It is established that for 2001–2024, the identified relationships have significantly strengthened, which indicates the activation of the positive reverse causality in question in the region. The tasks set in the work have been solved: the locations of the uluses (districts) of Yakutia have been determined, as well as the months for which the considered relationships are the strongest and most stable. It is also shown that the number of fires on the entire territory of Yakutia serves as a more informative predictor of the prognostic models of the studied process for its uluses (districts) than the burned area. The results of the study suggest that it is possible to use the results of monitoring forest fires in Yakutia to create forecasts for the upcoming year for the identified uluses (districts). This is particularly important for optimizing fire management strategies in a changing climate, as the months of May — July account for the peak of forest fires.

Introduction. The human factor is the cause of 70–80% of industrial accidents. This is the reason for scientific interest in this topic. Researchers are studying the issues of assessing occupational injury risks based on individual employee qualities. However, nonparametric methods are not used when analyzing the relationship between these qualities and hazardous incidents. At the same time, parametric statistical approaches for processing non-numeric information are unreasonable without first checking the distribution of variables for normality. The presented scientific work is intended to correct the situation. The aim is to identify and statistically substantiate the relationship between individual factors and realized production risks.

Materials and Methods. The authors observed the staff of Gazprom Transgaz Surgut LLC and created a questionnaire. They anonymously interviewed 569 workers and measured their level of responsibility (according to 34 statements) and emotionality (according to 26 statements). Eight variables were used in data processing: “injury”, “age”, “education”, “length of service in the company”, “total years of service”, “profession”, “responsibility”, and “emotionality”. The statements of 206 people (36.2%) with experience of injuries and occupational diseases and 363 (63.8%) without such experience were summarized. The correlation of independent variables and dependent variable (“injury”) was studied using the contingency tables. Estimates of the Pearson’s chi-square and the level of its statistical significance were supplemented by calculations of the intensity and direction of the relationship of variables (gamma coefficient).

Results. High internal consistency of the statements (Cronbach's alpha 0.923) and high content validity of the questionnaire have been proven. The significance (р < 0.001) for all variables allowed us to reject the null hypothesis about the subordination of the studied set of features to a normal distribution. Median frequencies of the “degree of responsibility” and “emotionality” variables for the group without injuries were 2 and 3, respectively, and noticeably higher in the group with injuries (5 and 5). The group differences in the statements were statistically significant (р < 0.05). Employees with realized risks had higher than average ranks in terms of qualitative severity of responsibility and emotionality. There was no significant difference in socio-demographic indicators for the grouping feature “injury” (р > 0.05). The Pearson’s chi-square value was 78.704 for the “responsibility — injury” pair and 35.350 for the “emotionality — injury” pair. Gamma was 0.514 and 0.359, respectively. Spearman's coefficient was 0.344 for responsibility and injury, 0.242 for emotionality and injury. The significance of all three criteria was <0.001.

Discussion. The outcome of risks was determined by individual rather than socio-demographic characteristics of the employees. This was indicated by:

• high median frequencies of “responsibility” and “emotionality” variables in the “injury” group,
• average ranks among respondents with realized hazardous events (р< 05).

The risk increased with increasing severity of signs of emotional instability and low responsibility. The relationship between “responsibility” and “injury” in gamma was stronger than in Spearman’s, therefore, gamma better accounted for nonlinear monotonic trends and showed a more significant monotonic average relationship.

Conclusion. Responsibility and emotionality are significant determinants of hazardous incidents. The research results will allow us to develop occupational safety and select staff more effectively. In the future, it is possible to build personalized (targeted) approaches to work with employees, and depending on their individual characteristics, predict the occurrence of hazardous events.

Introduction. Over the past decades, the issue of household and industrial waste management in Russia and worldwide has gained increased importance due to growing urbanization, increased consumption, and limited landfill space. The traditional disposal model — landfilling — is associated with environmental and public health risks and low levels of resource recovery. Alternative strategies are being developed, among which the key ones are separate collection and mechanical-biological treatment (MBT), as well as thermal disposal methods. A literature review shows that international studies (Europe, Japan) demonstrate the positive effects of a combination of separate collection and MBT, while Russian studies emphasize the barriers: insufficient infrastructure for separate collection, low public participation, and municipal financial constraints. The analysis highlights the importance of an integrated approach, where technological solutions are accompanied by organizational measures and economic incentives. Currently, there is a lack of comprehensive assessments of the MTB implementation, taking into account various separate collection and local logistics scenarios. Existing studies are often limited to technical and technological aspects or present calculations at the level of individual pilot sites, without assessing the systemic economic impacts or providing a detailed sensitivity analysis of key parameters. Therefore, the aim of this study is to identify optimal MSW sorting options for subsequent incineration, based on combustion heat and pollutant emissions.

Materials and Methods. The author collected and analyzed statistical data on the functioning of the MSW management system at the regional level using publicly available information from the Federal State Statistics Service and the Federal Service for Supervision of Natural Resources (Form 2-TP (waste))¹ as sources. Data and reports from IFC (International Finance Corporation)² and the public legal company Russian Environmental Operator (REO)³ were studied and analyzed. To achieve the set goals, the author has developed a research algorithm, including the use of regression analysis, the construction of a ranked series of the specific heat of MSW combustion by morphological composition with subsequent division into groups I and II, the formation of alternative waste sorting options by changing the proportion of MSW groups (I and II) in the conditionally accepted scenarios, followed by solving a multi-criteria problem of choosing the optimal option for MSW sorting for the purposes of further incineration. The research was conducted in MS Excel and Statistica software.

Results. An analysis of statistical data on the solid municipal waste management system and composition revealed that the morphological composition for the purposes of further incineration remained relatively stable over the past 5 years with minor seasonal fluctuations. A clear correlation was observed between population size and waste accumulation volume, as well as between waste volume and the level of waste sorting. Simulation modeling of a feasible set of alternative MSW sorting options made it possible to determine the optimal balance between the economic efficiency of incineration and environmental pollution. In this setting, batches were simulated for subsequent incineration with different ratios of solid municipal waste from groups I and II in the ratio of 10/90, 30/70, 50/50, 70/30, 90/10, respectively. Optimal options for MSW sorting for further incineration were determined by calculating a function that determined the degree of approximation of each of the sorting alternatives under consideration to the ideal value (ideal point), which was based on the values of minimum emissions (fly ash, sulfur oxide, nitrogen oxide, carbon monoxide) and the maximum value of the calorific value (the energy value of MSW based on the morphological composition). The sensitivity coefficients to changes in the basic values of the sorting parameters and the gross emission of pollutants were calculated. Optimal options for MSW sorting from the perspective of further incineration were obtained based on the calculation of combustion heat and pollutant emissions (including carbon footprint indicators).

Discussion. The main result of the study is a developed approach to the efficient sorting of municipal solid waste for subsequent thermal recycling. The solution to this problem was based on principles that allowed for a classical solution to the direct and dual problem of achieving maximum efficiency from the municipal solid waste incineration process while minimizing the negative impact on the environment. The obtained results will enable the effective separation and sorting of municipal solid waste despite the production constraints of incineration plants at municipal solid waste landfills. The analysis of the morphological composition of municipal solid waste for subsequent sorting and thermal recycling made it possible to identify waste separation options with the highest energy value based on heat of combustion of individual components and characterized by a minimum amount of pollutant emissions.

Conclusion. The analysis of statistical data from the municipal solid waste management system revealed the prevalence of landfill disposal and the ineffectiveness of the sorting process. Regression models revealed increasing trends in the volume of waste removed and sorted, necessitating further targeted and effective sorting for recycling purposes. Simulation modeling of alternative MSW sorting options and the solution of an optimization problem made it possible to identify effective MSW sorting options for further thermal disposal, with minimal pollutant emissions and maximum combustion heat values. In conclusion, during the research, optimal options for municipal solid waste sorting were obtained based on the morphological composition of batches for maximum efficiency of thermal disposal and minimum emissions of pollutants into the atmosphere.

Introduction. Providing safe drinking water to the population is a crucial task for health protection and sustainable development, as its quality directly influences the level of morbidity and mortality. The UN and WHO have stated that insufficient efficiency of water treatment systems contribute to the emergence and spread of infectious diseases, causing up to 1.4 million deaths worldwide each year. At the same time, the issue of comprehensive comparative assessment of water treatment technologies in terms of the overall risk to public health remains underdeveloped, considering both chemical and microbiological hazards. This gap in scientific knowledge necessitates research that focuses not only on meeting water quality standards but also on an integrated assessment of the effects of various technological schemes on human health. In this study, we aim to conduct a comparative assessment of the effectiveness of drinking water treatment technologies used in centralized water supply systems in terms of the overall risk to public health. This will made it possible to choose the best solution for water treatment in practice.

Materials and Methods. The information base for the study consisted of current regulatory documents that establish requirements for drinking water quality and technological processes for its preparation, such as the “Methodology for developing a register of BAT for water supply and sanitation systems”¹; Russian and international standards, and guidelines for assessing public health risks, scientific articles and monographs on filtration, coagulation, clarification, sorption, oxidation, and disinfection of water. The assessment of source water quality was conducted according to the main groups of indicators: organoleptic, generalized, sanitary-microbiological, parasitological, as well as sanitary-chemical. Mathematical modeling and statistical data processing methods were used to quantify and compare different water treatment schemes. The calculation was performed in accordance with the approaches described in MR 2.1.4.0289–22².

Based on the classification of water supply sources by water quality, we analyzed the recommended sets of technological operations:

• for the first class — pre-filtration with optional reagent treatment and mandatory disinfection;
• for the second class — filtration (in the presence of phytoplankton, microfiltration) with coagulation, settling and subsequent disinfection;
• for the third class — additional stage of purification including clarification, oxidation, sorption and repeated disinfection.

The study was performed using standard methods of laboratory analysis of water quality and specialized software for modeling and risk assessment.

Results. The effectiveness of the current treatment technology (mechanical purification, coagulation, and chlorination) and the proposed multistage scheme (including ultrafiltration, sorption, and combined disinfection) were evaluated. Mathematical modeling of changes in water quality parameters for three scenarios of water treatment was performed. Using special software, a model experiment and an assessment of quality changes were conducted for four groups of parameters (organoleptic, generalized, sanitary-microbiological and parasitological, and sanitary-chemical). According to MR 2.1.4.0289–22³, the values of integrated risk and the effectiveness of its reduction as a result of water treatment were calculated. The results were statistically processed. Based on the data on sanitary and hygienic monitoring and calculation of the overall risk to public health, the source water was found to have excesses in several indicators. It was established that the proposed multi-stage method provided more thorough purification and significantly reduced the negative impact on health across all groups of parameters (organoleptic, generalized, sanitary-microbiological and sanitary-chemical).

Discussion. A comparative analysis of the effectiveness of the two water treatment methods revealed a significant advantage of the multi-stage purification process. The proposed integrated approach fully ensured that water quality met the regulatory requirements for maximum permissible values through a combination of ultrafiltration, sorption and combined disinfection. The multi-stage purification scheme ensured not only complete microbiological and chemical safety, but also high organoleptic water parameters, enhancing the overall reliability of the water supply system.

Conclusion. The paper provides a comparative assessment of the effectiveness of two water treatment technologies for the centralized water supply system in Penza. Based on the methodology for calculating the overall risk to public health, it was found that the source water from the Surskoye reservoir had a high risk level. The current purification method (coagulation and chlorination) has been shown to reduce the risk to an average level, leaving the water supply system vulnerable. In contrast, the proposed multi-stage method (ultrafiltration, sorption, UV disinfection, and chloroamination) demonstrated very high efficiency (82%) in reducing the cumulative risk to negligible value. These results support the advantages of a multi-stage approach and can serve as a foundation for upgrading water treatment systems to increase their reliability and safety for the public.

Introduction. In industrial filtration systems, one of the main challenges is reducing the filter capacity due to the accumulation of retained particles and the formation of sediment layer on the filter baffle. This results in increased hydraulic resistance, increased energy consumption, and forced service stops. Extending the lifespan of filter elements while maintaining productivity is a crucial technological challenge. This involves methods such as the regeneration of hydrodynamic filters, including the rotation of the filter element and the use of vibration effects. However, current research focuses on these methods individually, with no theoretical models for the combined effect of centrifugal and vibrational forces. Experimental data on the synergy between these forces has not been collected, and criteria for optimizing this combined effect have not been established considering operating parameters and the adhesive properties of sediment. The aim of this research was to develop a computational method for optimizing the combined centrifugal-vibration effect, based on an analytical and experimental study of its impact on the regeneration efficiency of hydrodynamic filters.

Materials and Methods. The research was conducted on a laboratory test bench with a hydrodynamic vibrating filter equipped with a cylindrical filter baffle made of a combined porous mesh metal (fineness of 10 µm), which could perform independent rotational and vibrational movements. To describe the condition for sediment particle detachment, an analytical model was developed based on the balance of forces acting on a particle on a rotating and vibrating surface. This allowed us to evaluate the effectiveness of filter regeneration based on operating parameters. The experiments were conducted using aqueous suspensions of electrocorundum (200–250 µm) and silicon carbide (60–80 µm) with a volume concentration of 0.1%. The regeneration mode involved a simultaneous increase in the rotational speed of the baffle to 1000 rpm and vibration with an amplitude of 1 mm at a variable frequency of 50, 60 and 70 Hz with the filtrate outlet closed to eliminate the change of pressure.

Results. Quantitative dependencies of the regeneration efficiency on rotational speed, vibration amplitude and frequency were experimentally determined. An analytical model of force balance was developed, which allowed predicting the degree of purification for any combination of these parameters. Verification of the model showed that the discrepancy between the calculated and experimental data did not exceed 15–20%, confirming its suitability for engineering calculations. Based on the model, a computational optimization method was proposed that provided a choice of a combination of operating parameters at which the required level of cleaning was achieved with minimal energy consumption and permissible mechanical loads on the structure.

Discussion. The low efficiency of purely centrifugal regeneration (2–20%) was explained by the fact that for fine particles, the ratio of adhesive forces to inertial forces was significantly higher than for coarse particles. This was consistent with the Derjaguin classical theory of adhesion. The synergistic effect of the combined effect was due to the addition of radial centrifugal force by tangential shear stresses generated by vibration, which ensured a more complete destruction of adhesive bonds in the sediment layer. The discrepancy between the model and the experiment in the range of 15–20% was mainly due to uncertainty in determining the adhesion characteristics of the particle –filter baffle pair. However, this level of accuracy was acceptable for the engineering selection of operating parameters. The obtained patterns were qualitatively consistent with the known literature data on the individual effects of rotation and vibration on sediment removal, but for the first time, they quantitatively describe their combined effect. One limitation of the study was the validation of the model for aqueous suspensions only, which required additional research to extend it to viscous and non-Newtonian media.

Conclusion. It has been experimentally proven that the combination of centrifugal and vibrational effects can increase the regeneration efficiency of the hydrodynamic filter baffle by 60–80%, compared to 2–20% with rotation alone. An analytical model has been developed based on the balance of forces, and verified experimentally with an error of no more than 20%. This model is suitable for engineering calculations of optimal regeneration modes. It is demonstrated that the key parameter determining the accuracy of the forecast is the adhesion properties of particles, which require experimental determination for each system. The results provide a scientific basis for designing continuous self-cleaning filtration devices. A promising direction for future research is the adaptation of this technique to rheologically complex industrial environments, as well as optimizing energy consumption in the vibration system.

Introduction. In light of the increasing pollution of water resources by organic compounds, particularly synthetic dyes, it is a pressing issue to develop effective, affordable, and environmentally friendly sorption materials. Despite the widespread use of activated carbons, clays, and organomineral composites, there remains a need to find low-cost sorbents based on industrial waste. One promising approach is the use of clay sludge generated during the regeneration of machine oil, but their sorption properties have not been sufficiently studied. The aim of this research was to obtain and characterize the sorption characteristics of a material based on clay sludge waste during the removal of methylene blue from aqueous solutions.

Materials and Methods. The sorption material was obtained by thermal treatment of oily clay sludge at various temperatures, with the optimal mode selected. Pore structure was studied by low-temperature nitrogen adsorption using BET, t-Plot, and BJH models. Sorption properties were evaluated using model methylene blue solutions and photocolorimetry at a wavelength of 670 nm. Adsorption capacity and purification efficiency were calculated using standard methods, and sorption isotherms were approximated using the Langmuir, Freundlich, and Dubinin-Radushkevich models.

Results. It was found that the sample (CS400) heat-treated at 400°C had a developed mesoporous structure with a specific surface area of 69.148 m²/g and a total pore volume of 0.159 cm³/g. The average pore diameter was approximately 4–6 nm, with no micropores present. The material demonstrated high sorption activity for methylene blue, effectively decolorizing solutions. The maximum sorption capacity reached 0.139 mmol/g (44.8 mg/g). The sorption process was best described by the Langmuir model (R² = 0.9645), indicating monolayer nature of adsorption. The calculated sorption energy (9.608 kJ/mol) suggested a predominance of physical interaction.

Discussion. The results obtained demonstrated that the high sorption activity of the material was due to the formation of a mesoporous structure during heat treatment. Pores with a diameter of 4–6 nm were predominant, which ensured accessibility of the active surface to dye molecules. Hysteresis indicated the contribution of capillary condensation to the sorbate retention process. The compliance with the Langmuir model indicated relative homogeneity of active sites. The established physical nature of adsorption suggested the predominance of weak intermolecular interactions.

Conclusion. The feasibility of the effective use of thermally modified clay sludge waste as a sorbent for the purification of water from cationic dyes has been experimentally confirmed. The CS400 material has been shown to have a high sorption capacity, effectively removing methylene blue from aqueous solutions. The obtained results demonstrated the potential of the developed sorbent for use in water treatment technologies and highlighted the feasibility of recycling industrial waste to produce functional materials.

Introduction. Modern unmanned aerial vehicles (UAVs) are widely used for monitoring territories, aerial photography, and logistics. Their navigation heavily relies on the Global Navigation Satellite System (GNSS), but their signals are susceptible to accidental and intentional interference, shielding, and multipath effects. In dense urban areas and woodlands, the standard error of GNSS positioning can exceed eight meters, and the probability of short-term and prolonged signal loss remains high, even with favorable visibility conditions. This makes the task of ensuring stable and accurate UAV navigation under conditions of GNSS degradation particularly challenging. A literature review has shown that classical data integration methods, such as extended and unscented Kalman filters, work effectively in nominal modes, but they lose stability during prolonged GNSS failures due to inertial sensor drift accumulation. New architectures based on deep learning (e.g. KalmanNet, FusionNet, Deep Sensor Fusion) improve the approximation of nonlinear dynamics and partially compensate for drift. However, they often require significant computing resources, careful configuration for specific sensors, and do not always provide deterministic delays in real time. Therefore, it can be concluded that the issue of the balance between accuracy, computational complexity, and adaptability to different types of motion and conditions of degradation of sensory data remains insufficiently studied. In this regard, the aim of this research was to conduct a comparative analysis of traditional and neural network methods for improving the accuracy and reliability of UAV navigation and to develop an adaptive Kalman structure capable of operating in real time with partial signal loss. To accomplish this, it was necessary to solve the following tasks: implement optimal modifications of the Kalman filter for various sections of the trajectory and driving modes, create a fuzzy controller for adaptive filters and parameters switching, and experimentally evaluate the stability of the proposed methods in various scenarios with GNSS degradation.

Materials and Methods. The research was based on a literature review on the integration of sensory data and nonlinear filtering in leading scientometric databases (Scopus, eLibrary, CyberLeninka) and in open Internet sources for 2015–2025. Mathematical modeling was conducted in the MATLAB environment. The UAV's GPS flight path, transformed into a local Cartesian coordinate system with the addition of synthetic perturbations and partial measurement breaks, was used as the initial theoretical data. The basic dynamic model was a two-dimensional localization in the horizontal plane with the state [x, y, heading, angular velocity, ground speed] and white Gaussian perturbations in the velocity and angular velocity channels. The Euler method was used for discretization. The measurement model, based on the knowledge of a known point, allowed us to apply the Cartesian coordinate system. EKF, UKF/SRCDKF and Particle Filter were implemented and compared as reference algorithms for nonlinear filtering. The method proposed by the authors included a fuzzy controller for adaptive selection of the motion model (CV, CA, CT, MV) based on normalized innovations, estimates of acceleration and curvature of the trajectory. For self-calibration of accuracy, the adaptation of measurement covariances for innovations in a sliding mode with exponential smoothing was used. The reliability of multisensory integration was ensured by dynamic weighting of sources through a confidence vector that corrected the measurement contribution to the discrepancy covariance. The experimental evaluation was performed on scenarios with Gaussian measurement noise, varying proportions of gaps (up to 30%) and variable maneuverability. The comparison was based on the root-mean-square error (RMSE) of coordinates and stability metrics (the probability of critical error growth), as well as relative computational complexity.

Results. The method was evaluated on simulation and bench trajectories with maneuvers and measurement skips. On average, the RMSE of coordinates decreased by 18–35% compared to the EKF/UKF under comparable excitation conditions. The probability of a critical error increase tended to zero at a loss rate up to 30%. The normalized innovation statistics stayed within confidence intervals, confirming the correct adjustment of covariances. Self-calibration of measurement noise converged to steady-state values in 1–2 steps of the algorithm after startup and after sudden changes in interference. Ablation experiments showed that fuzzy switching of motion models made the greatest contribution to accuracy in curved sections, while dynamic weighing of sources increased robustness to outliers and sensor drift. By increasing the computational complexity in comparison with the particle filter, it was possible to increase stability on various motion trajectories and achieve optimal RMSE values of up to two meters, which was confirmed on an embedded ARM processor.

Discussion. The gain in accuracy and stability was due to a combination of a locally adequate kinematic model and online adaptation of sensor confidence, which reduced systematic biases and prevented covariance overclocking. Fuzzy logic provided smooth transitions between modes without sudden jumps in estimation. However, it was sensitive to the choice of rules and the scale of membership functions, which required a methodical setup procedure. Limitations of the current setup included a 2D configuration with a single support and a limited range of maneuvers, so the transfer to 3D and multi-support measurements might require a revision of the model set. Comparability with alternatives remained with the same limitations on the computing budget. With unlimited resources, heavier methods partially reduced the gap. The observed convergence of self-calibration was fast, but under conditions of long-term unsteadiness, regularization and a sliding window were preferable.

Conclusion. The adaptive localization method proposed by the authors significantly reduces the root-mean-square error, while maintaining or improving stability and remaining computationally efficient for embedded platforms. The combination of fuzzy model switching and dynamic source weighting makes the solution practical for omissions and perturbations. Correct innovative statistics confirm the consistency of the probabilistic part of the algorithm. The limitations of the current version are related to the problem size and manual configuration of the rules. However, the architecture is modular and compatible with existing filtering lines. Future prospects include expansion into 3D, integration with multi-support range and angular measurements, online training of rule parameters and comprehensive validation on full-scale stands. Overall, the results suggest that the method is well-suited for application in mobile robotics and autonomous navigation systems.

Introduction. The use of artificial neural networks (ANNs) to diagnose the technical condition of automotive equipment is an active area of research. However, existing work mainly focuses on evaluating individual units, such as the engine, without a comprehensive analysis of the interconnected systems of a car. This creates a gap in the field of the development of intelligent systems that can take into account the state of the chassis, braking, and steering systems at the same time. The aim of this study is to develop an intelligent decision-making support system (IDMSS) based on ANNs that can comprehensively assess the technical condition of a vehicle by combining expert knowledge and data on damage to different components.

Materials and Methods. Defective indicators, determined on the basis of regulatory documents, and manuals on operation, maintenance and repair, were used to defect car parts and assemblies. The research was based on the methodology of neural network modeling. To train the ANN, an array of 100 samples was used, formed on the basis of:

• statistical data;
• expert surveys of specialists from the Automotive Equipment Maintenance and Repair Center at Don State Technical University;
• analysis of big data from online sources.

Defective parameters of 13 main vehicle systems, operational factors and even the psycho-emotional state of the driver were considered. The training array included damage parameters for frame parts, axles, suspension, wheels, brake, and steering systems. To compare the effectiveness, three multilayer perceptrons (MLPs) architectures with different numbers of neurons in hidden layers, activation functions, and the BFGS learning algorithm were created and trained.

Results. The best results were shown by the MLP 8-24-3 neural network (8 input, 24 hidden, 3 output neurons). Its performance on the training sample was 93.75%, on the test sample — 90%. The accuracy of classification by category of technical condition reached 100% for the category “operation permitted”, 94.74% for “operation permitted with restrictions”, and 82.35% for “operation prohibited”. Sensitivity analysis revealed that the parameters of the frame (X1) and axles (X2) had the greatest influence on the classification.

Discussion. The developed ANN has demonstrated high efficiency in a comprehensive assessment of the vehicle's technical condition, going beyond the diagnosis of individual units. It has been established that the weighting coefficients of the neural network can serve as a quantitative measure of the relationship and mutual influence of the details of various systems on the overall safety. The results obtained confirm the practical applicability of the approach for creating flexible IDMSSs in the field of maintenance and diagnostics.

Conclusion. The research contributes to the development of data mining methods for transport systems, offering a new approach to integrating heterogeneous parameters and expertise into a single neural network model. It is an important step towards improving the reliability and safety of automotive equipment. An intelligent system based on expert experience and statistical data is a promising tool for automating assessment and decision-making processes. Further development of the system may include expanding the database and improving learning algorithms, which will increase its accuracy and efficiency.

Introduction. Pollution of aquatic ecosystems by petroleum products, including the transboundary transport of pollutants from ships' ballast water, requires improvement of cleaning methods. Existing shipboard ballast water management systems are not sufficiently effective in removing dissolved and emulsified hydrocarbons. A promising solution is the use of sorption materials. However, choosing the optimal sorbent for specific pollutants is a challenging task that requires scientific research. In this study, we aimed to demonstrate a quantum chemical modeling technique to predict the effectiveness of the “sorbent — pollutant” interaction using cellulose and typical oil components as examples.

Materials and Methods. A fragment of cellulose (cellobiose) and contaminant molecules: benzene, phenol, and naphthalene were used as a model system. These substances were chosen due to their chemical structure and ability to simulate real environmental pollution. Preliminary optimization of the geometry and calculation of energy parameters were performed using the semi-empirical PM3 method in the GAMESS program. To verify the results, the density functional theory with the B3LYP functional and the 6-31G(d) basis was used. The adsorption energy was calculated as the difference between the total energies of the complex and the isolated components. The active interaction centers were identified based on the analysis of geometric parameters, boundary molecular orbitals (HOMO/LUMO), and charge transfer.

Results. The key electronic characteristics of pollutants were calculated, showing that naphthalene had the highest polarizability (HOMO-LUMO gap 8.43 eV), and phenol had a significant dipole moment (1.14 D). Geometrically and energetically optimal configurations were determined for the cellobiose-benzene complex. It was established that sorption was provided by the formation of weak hydrogen bonds (O...H-C) with distances of 1.85-1.91 Å. The adsorption energy for the most stable configuration was 21.27 kJ/mol, which corresponded to a stable non-covalent interaction. Criteria for the stability of adsorption complexes (energy, structural, electronic) were formulated for the development of preliminary heuristic rules in the decision support system for the selection of sorbents.

Discussion. The developed quantum chemical modeling technique made it possible to quantify the energy and mechanisms of intermolecular interaction in the "sorbent — pollutant" system. It was shown that native cellulose was able to effectively retain nonpolar aromatic hydrocarbons due to dispersion forces and weak hydrogen bonds. The calculated parameters can serve as the basis for a scientifically sound selection of components for ballast water filters and other purification systems, taking into account the type of pollutant, as well as for integration into information and analytical decision support systems.

Conclusion. The results of the work can be integrated into information and analytical decision support systems for the selection of sorbents for ballast water treatment, as well as serve as a basis for further research of modified forms of cellulose.