High-speed railway power quality evaluation system and method based on three-layer fusion weighting

Abstract

The invention relates to the technical field of power quality management, in particular to a high-speed railway power quality evaluation system and method based on three-layer fusion weighting, and the system comprises a game theory dynamic weight mechanism, a fuzzy VIKOR evaluation theory, gray correlation analysis and a variance-based adaptive fusion mechanism. According to the high-speed railway power quality evaluation system and method based on three-layer fusion weighting, the three-phase voltage unbalance degree and the voltage deviation are objectively identified as key control indexes through a game theory dynamic weight mechanism, and the uncertainty of the system is effectively processed through fuzzy VIKOR evaluation; the internal relation between the measuring points is deeply mined through grey correlation analysis, advantage complementation of multiple methods is achieved through an adaptive fusion mechanism, finally, scientific and reasonable evaluation sorting is formed, and the real electric energy quality level difference of all the measuring points is accurately reflected. In future research, an intelligent real-time evaluation system can be constructed in combination with an artificial intelligence technology, and dynamic monitoring and early warning of the electric energy quality of the high-speed railway system are realized.

Description

Technical Field

[0001] This invention relates to the field of power quality management technology, specifically to a high-speed railway power quality assessment system and method based on three-layer fusion weighting. Background Technology

[0002] As a crucial infrastructure for modern transportation, high-speed railways place extremely stringent demands on power quality for safe and stable operation. The traction system of high-speed trains employs high-power converters and high-frequency switching technology, generating complex harmonic interference and voltage fluctuations during operation. Furthermore, the widespread application of regenerative braking systems further exacerbates the complexity of power quality issues. Unlike traditional power systems, high-speed railway power supply systems are characterized by high load mobility, drastic power fluctuations, and high sensitivity to power quality. Any power quality problem can directly impact train safety and passenger experience. Therefore, establishing a scientific, objective, and comprehensive power quality assessment system for high-speed railways is of great significance for ensuring railway transportation safety, improving service quality, and guiding system optimization.

[0003] Traditional power quality assessment methods have revealed several limitations when applied to high-speed railway systems. First, regarding weight determination, traditional subjective methods such as the analytic hierarchy process (AHP) struggle to accurately reflect the true importance of each power quality indicator under the dynamic operating environment of high-speed railways, particularly failing to effectively handle the impact of rapid changes in traction load on indicator weight allocation. Second, in dealing with uncertainties, high-speed railway power quality data exhibits significant randomness and fuzziness, making it difficult for traditional deterministic assessment methods to fully consider the impact of measurement errors, environmental interference, and equipment aging on the assessment results. Furthermore, in analyzing indicator correlations, existing methods often treat each power quality indicator as an independent variable, neglecting the complex coupling relationships and dynamic interactions between indicators within the high-speed railway traction system. Summary of the Invention

[0004] To address the shortcomings of existing technologies, this invention provides a high-speed railway power quality assessment system and method based on three-layer fusion weighting, which constructs an innovative "three-layer fusion" assessment architecture suitable for high-speed railway power quality assessment.

[0005] To achieve the above objectives, the following technical solution is provided:

[0006] A high-speed railway power quality assessment system based on three-layer fusion weighting is characterized by comprising a game theory dynamic weighting mechanism, fuzzy VIKOR evaluation theory, grey relational analysis, and a variance-based adaptive fusion mechanism; wherein,

[0007] The game theory dynamic weighting mechanism is the basic layer. By treating each power quality indicator as a rational game subject, it automatically identifies the key indicators with the strongest distinguishing ability in the high-speed railway operating environment and realizes the objective dynamic allocation of weights.

[0008] The fuzzy VIKOR evaluation theory is the core layer. The triangular fuzzy number theory is used to model the uncertainty in the power quality measurement process of high-speed railway, and effectively handle the evaluation fuzziness caused by traction load fluctuations and changes in environmental factors.

[0009] The grey relational analysis is a supplementary layer that deeply explores the correlation between each measuring point and the ideal operating state, and fully considers the grey system characteristics of the high-speed railway power supply system.

[0010] The variance-based adaptive fusion mechanism enables collaborative decision-making among multiple methods, providing a basis for the assessment and systematic improvement of power quality in high-speed railways.

[0011] Preferably, in the game theory dynamic weighting mechanism, considering the dynamic characteristics and interrelationships of power quality indicators in the high-speed railway traction power supply system, m power quality indicators are regarded as rational game participants, and the payoff function of each indicator reflects its importance under the changing traction load environment:

[0012] (1);

[0013] In the formula Let w be the revenue function value of the i-th index in the traction power supply system. i w is the weight of the i-th indicator. j Let σ be the weight of the j-th indicator. i ² Let ρ be the variance of the i-th index under different traction load conditions. ij The correlation coefficient between indicators in the traction power supply system, and the adjustment parameter α i =1、β i =0.5、γ i =0.1 controls the impact of discrimination gain, relevance competition, and weight penalty, respectively;

[0014] The design of this revenue function takes into account the impact of rapid changes in the traction power of high-speed trains on the importance of various power quality indicators, and achieves automatic weight allocation that adapts to the dynamic characteristics of the traction power supply system through Nash equilibrium solution.

[0015] Preferably, in the fuzzy VIKOR evaluation theory, considering the multiple uncertainties in the power quality measurement process of high-speed railway traction power supply systems, such as traction load fluctuations, dynamic contact between the contact wire and pantograph, and changes in environmental factors, the traditional VIKOR method needs to be improved by fuzzification to adapt to the complex operating environment of the traction power supply system; for the j-th power quality index at the i-th measurement point, a triangular fuzzy number is constructed:

[0016] (2);

[0017] Among them, F̃ ij To account for the triangular fuzzy number of the traction power supply system uncertainty, x ij The standardized index value, σ j This represents the standard deviation of the index under different traction conditions.

[0018] The construction of fuzzy numbers fully considers the impact of different operating conditions of high-speed trains—starting, braking, and constant-speed operation—on power quality. The group utility S is calculated using fuzzy distance. k and individual regrets R k The VIKOR comprehensive index of the k-th measurement point is obtained as follows:

[0019] (3);

[0020] This formula balances the two dimensions of group utility and individual regret in the evaluation of traction power supply systems. It considers both the overall power quality level of each traction substation and the prominent problems of individual substations in specific indicators, where S * S — R represents the minimum and maximum group utility, respectively. * R — These represent the minimum and maximum values ​​of individual regrets, respectively. This balancing mechanism is particularly suitable for assessment needs in traction power supply systems that require both ensuring overall power supply quality and preventing individual components from becoming limiting factors.

[0021] Preferably, in grey relational analysis, the high-speed railway traction power supply system, as a typical grey system, is affected by various factors such as train timetable, traction load distribution, and equipment status, exhibiting characteristics that are neither completely known nor completely unknown. The variation law of power quality in the traction power supply system is often difficult to describe precisely using deterministic mathematical models, while grey relational analysis can effectively handle this system characteristic of incomplete information. Grey relational analysis is used to explore the deep correlation between each traction substation and the ideal power supply state. The formula for calculating the correlation coefficient of the j-th power quality index at the k-th measuring point is:

[0022] (4);

[0023] In the formula, ξ k (j) is the correlation coefficient between the traction substation and the ideal power supply state, Δ k (j) represents the absolute difference between the j-th index of the k-th traction substation and the ideal value, Δ min and Δ max These are the minimum and maximum values ​​of all absolute differences, respectively;

[0024] The weighted correlation degree of the k-th measurement point is obtained by weighted summation based on the weights determined by game theory:

[0025] (5).

[0026] In the formula, The weighted correlation degree of the k-th measurement point. The weight of the j-th indicator is determined by game theory.

[0027] Preferably, in the variance-based adaptive weight fusion mechanism, to fully leverage the different advantages of fuzzy VIKOR and grey relational analysis in power quality assessment of traction power supply systems and improve the reliability of results, a variance-based adaptive weight fusion mechanism is designed. This mechanism considers the diversity and complexity of the operating conditions of the traction power supply system, and judges its reliability by evaluating the dispersion of the results of each method under different traction load conditions. The larger the variance, the stronger the discrimination ability of the method under the current traction power supply system state. The final comprehensive score of the k-th measurement point is:

[0028] (6);

[0029] In the formula, This represents the final comprehensive score for the k-th measurement point. The fuzzy VIKOR score of the kth measurement point The weights are the weights of the fuzzy VIKOR score for the k-th measurement point, obtained through a variance-based adaptive weight fusion mechanism. The grey relational analysis score of the k-th measurement point The weights of the grey relational analysis score of the k-th measurement point are obtained based on the variance-adaptive weight fusion mechanism.

[0030] This adaptive mechanism ensures that the fusion evaluation results can reflect the power quality level of each traction substation under different traction load distributions and operating conditions.

[0031] The high-speed railway power quality assessment system based on three-layer fusion weighting includes the following steps:

[0032] Step 1: Data preprocessing. Load the power quality monitoring data of each measuring point, including seven core indicators: voltage fluctuation, three-phase voltage imbalance, voltage deviation, total harmonic distortion rate, odd voltage harmonics, even voltage harmonics, and frequency deviation. Clean the raw data, remove outliers and fill in missing data, and then perform normalization and standardization to map all indicators to the 0-1 range, laying the data foundation for subsequent evaluation and analysis.

[0033] Step 2: Game theory dynamic weight allocation. The seven power quality indicators are regarded as rational game participants. A payoff function is constructed that considers the indicator's distinguishing ability, correlation competition and weight penalty. The objective weight of each indicator is solved by Nash equilibrium optimization algorithm to automatically identify the key indicator with the most distinguishing ability in the current data environment.

[0034] Step 3: Fuzzy VIKOR evaluation. Construct triangular fuzzy numbers to handle uncertainties in the evaluation process, convert standardized index values ​​into fuzzy number form to account for measurement errors and environmental interference; determine positive and negative ideal solutions as reference benchmarks, calculate the group utility value and individual regret value of each measurement point, and finally generate a VIKOR comprehensive index that balances multi-objective conflicts to obtain the evaluation results based on fuzzy theory.

[0035] Step 4: Grey relational analysis. Construct an ideal power quality state as a reference sequence, calculate the grey relational coefficient between each measuring point and the reference sequence on each index; use the weights determined by game theory to perform a weighted summation of the relational coefficients to obtain the comprehensive relational degree, and quantify the similarity and correlation between each measuring point and the ideal state from the perspective of grey system theory.

[0036] Step 5: Adaptive fusion evaluation. Analyze the evaluation stability of the two methods, fuzzy VIKOR and grey relational analysis, and evaluate the reliability of each method through the variance index. Based on the reliability, dynamically allocate fusion weights, and combine the evaluation results of the two methods in a weighted manner to generate the final comprehensive evaluation score and ranking, giving full play to the advantages of each method and avoiding the limitations of a single method.

[0037] Step 6: Results analysis. Based on the integrated evaluation scores, each measuring point is ranked and classified into levels, and the results are compared and verified with the national power quality standards. Based on the evaluation results, the main problems and weaknesses of each measuring point are identified, and the power quality evaluation is finally completed.

[0038] The beneficial effects of this invention are as follows: The high-speed railway power quality assessment system and method based on three-layer fusion weighting objectively identify three-phase voltage imbalance and voltage deviation as key control indicators through a game theory dynamic weighting mechanism; effectively handle system uncertainties through fuzzy VIKOR evaluation; deeply explore the intrinsic relationships between measurement points through grey relational analysis; and achieve complementary advantages of multiple methods through an adaptive fusion mechanism, ultimately forming a scientific and reasonable assessment ranking that accurately reflects the differences in the actual power quality levels of each measurement point. Future research can combine artificial intelligence technology to build an intelligent real-time assessment system to realize dynamic monitoring and early warning of power quality in high-speed railway systems. Attached Figure Description

[0039] Figure 1 This invention provides a power quality assessment index system.

[0040] Figure 2 This is a flowchart of the comprehensive power quality assessment process of the present invention;

[0041] Figure 3 This is a main wiring diagram of a joint-construction facility in an embodiment of the present invention; Detailed Implementation

[0042] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0043] The following detailed description of the construction process of the gene editing system of the present invention, with reference to the accompanying drawings, and further illustrates the usage method and experimental results of the gene editing system of the present invention through the examples.

[0044] Example 1

[0045] This invention establishes a power quality assessment index system from three dimensions: voltage quality, harmonic quality, and frequency quality. Figure 1 As shown in the figure. Among them, voltage quality considers the voltage fluctuation, three-phase voltage imbalance, and voltage deviation of the traction system; harmonic quality includes the total harmonic distortion and harmonic content of the voltage; and frequency quality considers the frequency deviation of the traction system.

[0046] According to relevant national standards, power quality indicators are divided into five levels. The values ​​of individual indicators in each power quality level are shown in Table 1. In the table and below, r1, r2, ..., r7 represent voltage fluctuation, three-phase voltage imbalance, voltage deviation, total harmonic distortion rate, odd voltage harmonic content, even voltage harmonic content, and frequency deviation, respectively.

[0047] Table 1. Classification of Power Quality Indicators

[0048] index Ⅰ Ⅱ Ⅲ Ⅳ Ⅴ <![CDATA[r1]]> ≤0.5 ≤1 ≤1.5 ≤2 >2 <![CDATA[r2]]> ≤0.5 ≤1 ≤1.5 ≤2 >2 <![CDATA[r3]]> ≤1.2 ≤3 ≤4.5 ≤7 >7 <![CDATA[r4]]> ≤1 ≤2 ≤3 ≤5 >5 <![CDATA[r5]]> ≤0.4 ≤0.8 ≤1.2 ≤1.6 >1.6 <![CDATA[r6]]> ≤0.2 ≤0.4 ≤0.6 ≤0.8 >0.8 <![CDATA[r7]]> ≤0.05 ≤0.1 ≤0.15 ≤0.2 >0.2

[0049] Example 2

[0050] A high-speed railway power quality assessment system based on three-layer fusion weighting is proposed. This system includes a game-theoretic dynamic weighting mechanism, fuzzy VIKOR evaluation theory, grey relational analysis, and a variance-based adaptive fusion mechanism.

[0051] The game theory dynamic weighting mechanism is the basic layer. By treating each power quality indicator as a rational game subject, it automatically identifies the key indicators with the strongest distinguishing ability in the high-speed railway operating environment and realizes the objective dynamic allocation of weights.

[0052] The fuzzy VIKOR evaluation theory is the core layer. The triangular fuzzy number theory is used to model the uncertainty in the power quality measurement process of high-speed railway, and effectively handle the evaluation fuzziness caused by traction load fluctuations and changes in environmental factors.

[0053] The grey relational analysis is a supplementary layer that deeply explores the correlation between each measuring point and the ideal operating state, and fully considers the grey system characteristics of the high-speed railway power supply system.

[0054] The variance-based adaptive fusion mechanism enables collaborative decision-making among multiple methods, providing a basis for the assessment and systematic improvement of power quality in high-speed railways.

[0055] Preferably, in the game theory dynamic weighting mechanism, considering the dynamic characteristics and interrelationships of power quality indicators in the high-speed railway traction power supply system, m power quality indicators are regarded as rational game participants, and the payoff function of each indicator reflects its importance under the changing traction load environment:

[0056] (1);

[0057] In the formula Let w be the revenue function value of the i-th index in the traction power supply system. i w is the weight of the i-th indicator. j Let σ be the weight of the j-th indicator. i ² Let ρ be the variance of the i-th index under different traction load conditions. ij The correlation coefficient between indicators in the traction power supply system, and the adjustment parameter α i =1、β i =0.5、γ i =0.1 controls the impact of discrimination gain, relevance competition, and weight penalty, respectively;

[0058] The design of this revenue function takes into account the impact of rapid changes in the traction power of high-speed trains on the importance of various power quality indicators, and achieves automatic weight allocation that adapts to the dynamic characteristics of the traction power supply system through Nash equilibrium solution.

[0059] Preferably, in the fuzzy VIKOR evaluation theory, considering the multiple uncertainties in the power quality measurement process of high-speed railway traction power supply systems, such as traction load fluctuations, dynamic contact between the contact wire and pantograph, and changes in environmental factors, the traditional VIKOR method needs to be improved by fuzzification to adapt to the complex operating environment of the traction power supply system; for the j-th power quality index at the i-th measurement point, a triangular fuzzy number is constructed:

[0060] (2);

[0061] Among them, F̃ ij To account for the triangular fuzzy number of the traction power supply system uncertainty, x ij The standardized index value, σ j This represents the standard deviation of the index under different traction conditions.

[0062] The construction of fuzzy numbers fully considers the impact of different operating conditions of high-speed trains—starting, braking, and constant-speed operation—on power quality. The group utility S is calculated using fuzzy distance. k and individual regrets R k The VIKOR comprehensive index of the k-th measurement point is obtained as follows:

[0063] (3);

[0064] This formula balances the two dimensions of group utility and individual regret in the evaluation of traction power supply systems. It considers both the overall power quality level of each traction substation and the prominent problems of individual substations in specific indicators, where S * S — R represents the minimum and maximum group utility, respectively. * R — These represent the minimum and maximum values ​​of individual regrets, respectively. This balancing mechanism is particularly suitable for assessment needs in traction power supply systems that require both ensuring overall power supply quality and preventing individual components from becoming limiting factors.

[0065] Preferably, in grey relational analysis, the high-speed railway traction power supply system, as a typical grey system, is affected by various factors such as train timetable, traction load distribution, and equipment status, exhibiting characteristics that are neither completely known nor completely unknown. The variation law of power quality in the traction power supply system is often difficult to describe precisely using deterministic mathematical models, while grey relational analysis can effectively handle this system characteristic of incomplete information. Grey relational analysis is used to explore the deep correlation between each traction substation and the ideal power supply state. The formula for calculating the correlation coefficient of the j-th power quality index at the k-th measuring point is:

[0066] (4);

[0067] In the formula, ξ k (j) is the correlation coefficient between the traction substation and the ideal power supply state, Δ k (j) represents the absolute difference between the j-th index of the k-th traction substation and the ideal value, Δ min and Δ max These are the minimum and maximum values ​​of all absolute differences, respectively;

[0068] The weighted correlation degree of the k-th measurement point is obtained by weighted summation based on the weights determined by game theory:

[0069] (5).

[0070] In the formula, The weighted correlation degree of the k-th measurement point. The weight of the j-th indicator is determined by game theory.

[0071] Preferably, in the variance-based adaptive weight fusion mechanism, to fully leverage the different advantages of fuzzy VIKOR and grey relational analysis in power quality assessment of traction power supply systems and improve the reliability of results, a variance-based adaptive weight fusion mechanism is designed. This mechanism considers the diversity and complexity of the operating conditions of the traction power supply system, and judges its reliability by evaluating the dispersion of the results of each method under different traction load conditions. The larger the variance, the stronger the discrimination ability of the method under the current traction power supply system state. The final comprehensive score of the k-th measurement point is:

[0072] (6);

[0073] In the formula, This represents the final comprehensive score for the k-th measurement point. The fuzzy VIKOR score of the kth measurement point The weights are the weights of the fuzzy VIKOR score for the k-th measurement point, obtained through a variance-based adaptive weight fusion mechanism. The grey relational analysis score of the k-th measurement point The weights of the grey relational analysis score of the k-th measurement point are obtained based on the variance-adaptive weight fusion mechanism.

[0074] This adaptive mechanism ensures that the fusion evaluation results can reflect the power quality level of each traction substation under different traction load distributions and operating conditions.

[0075] A high-speed railway power quality assessment system based on three-layer fusion weighting, such as Figure 2 As shown, it includes the following steps:

[0076] Step 1: Data preprocessing. Load the power quality monitoring data of each measuring point, including seven core indicators: voltage fluctuation, three-phase voltage imbalance, voltage deviation, total harmonic distortion rate, odd voltage harmonics, even voltage harmonics, and frequency deviation. Clean the raw data, remove outliers and fill in missing data, and then perform normalization and standardization to map all indicators to the 0-1 range, laying the data foundation for subsequent evaluation and analysis.

[0077] Step 2: Game theory dynamic weight allocation. The seven power quality indicators are regarded as rational game participants. A payoff function is constructed that considers the indicator's distinguishing ability, correlation competition and weight penalty. The objective weight of each indicator is solved by Nash equilibrium optimization algorithm to automatically identify the key indicator with the most distinguishing ability in the current data environment.

[0078] Step 3: Fuzzy VIKOR evaluation. Construct triangular fuzzy numbers to handle uncertainties in the evaluation process, convert standardized index values ​​into fuzzy number form to account for measurement errors and environmental interference; determine positive and negative ideal solutions as reference benchmarks, calculate the group utility value and individual regret value of each measurement point, and finally generate a VIKOR comprehensive index that balances multi-objective conflicts to obtain the evaluation results based on fuzzy theory.

[0079] Step 4: Grey relational analysis. Construct an ideal power quality state as a reference sequence, calculate the grey relational coefficient between each measuring point and the reference sequence on each index; use the weights determined by game theory to perform a weighted summation of the relational coefficients to obtain the comprehensive relational degree, and quantify the similarity and correlation between each measuring point and the ideal state from the perspective of grey system theory.

[0080] Step 5: Adaptive fusion evaluation. Analyze the evaluation stability of the two methods, fuzzy VIKOR and grey relational analysis, and evaluate the reliability of each method through the variance index. Based on the reliability, dynamically allocate fusion weights, and combine the evaluation results of the two methods in a weighted manner to generate the final comprehensive evaluation score and ranking, giving full play to the advantages of each method and avoiding the limitations of a single method.

[0081] Step 6: Results analysis. Based on the integrated evaluation scores, each measuring point is ranked and classified into levels, and the results are compared and verified with the national power quality standards. Based on the evaluation results, the main problems and weaknesses of each measuring point are identified, and the power quality evaluation is finally completed.

[0082] Example 3: Case Study

[0083] 1. Data Introduction

[0084] The measured data in this embodiment comes from five measuring points at a joint construction site of a high-speed railway. The locations and voltage levels of the measuring points are as follows: Figure 3 As shown in Table 2, statistical calculations were performed on the actual measurement data from five measuring points to obtain the measured data for seven power quality indicators: voltage fluctuation, three-phase voltage imbalance, voltage deviation, total harmonic distortion (THD), harmonic content, and frequency deviation. The 95% probability values ​​of the measured data were used as the evaluation data. The measured data show that the 25th and 30th voltage harmonics have relatively higher contents than other harmonics at the five measuring points. Therefore, odd-order voltage harmonics are represented by the 25th harmonic, and even-order voltage harmonics are represented by the 30th harmonic.

[0085] Table 2 Raw power quality data at each measuring point

[0086] measuring point <![CDATA[r1]]> <![CDATA[r2]]> <![CDATA[r3]]> <![CDATA[r3]]> <![CDATA[r4]]> <![CDATA[r5]]> <![CDATA[r6]]> 1 0.075 4.54 0.056 1.46 0.22 0.08 0.001 2 0.057 5.00 0.053 4.60 1.28 0.35 0.001 3 0.010 1.00 0.010 8.57 1.28 0.27 0.002 4 0.050 1.00 0.010 7.10 4.18 0.94 0.002 5 0.050 1.00 0.010 4.50 2.00 0.80 0.002

[0087] 2. Power quality assessment results

[0088] Step 1: The data preprocessing results are shown in Table 3:

[0089] Table 3. Preprocessed data for each measuring point

[0090] measuring point <![CDATA[r1]]> <![CDATA[r2]]> <![CDATA[r3]]> <![CDATA[r3]]> <![CDATA[r4]]> <![CDATA[r5]]> <![CDATA[r6]]> 1 0 0 0 0.66 0 0 0.207 2 0.485 0.47 0.173 0.00 0.42 0.49 0.429 3 0.997 0.88 0.962 0.91 0.91 0.91 0.207 4 0.741 0.61 0.771 0.66 0.67 0.76 0 5 1 1 1 1 1 1 1

[0091] Step Two: The game theory-based dynamic weight allocation mechanism successfully achieved the objective determination of the weights for power quality indicators. The final weight allocation results show that three-phase voltage imbalance received the highest weight of 0.30, reflecting the significant impact of single-phase traction loads on the balance of the three-phase power system. Voltage deviation also received a weight of 0.30, reflecting the stringent requirements for voltage stability imposed by the high power and rapid changes of traction loads. Odd-order voltage harmonics received a weight of 0.20, highlighting the severity of harmonic pollution problems in modern traction converters. The weights of the remaining indicators were all around 0.05, indicating that their distinguishing ability was relatively weak under the current data.

[0092] Step 3: The fuzzy VIKOR score shows that test point 3 achieved the highest score of 0.516, followed closely by test point 5 with 0.522. These two test points demonstrated relatively excellent performance in the fuzzy VIKOR evaluation. Test point 1 scored only 0.116, reflecting that its power quality problems were the most severe under uncertain conditions.

[0093] Step 4: The calculation results of the grey relational coefficient show that the degree of correlation between each measuring point and the ideal state varies significantly across different indicators. The choice of a resolution coefficient ρ=0.5 ensures the stability of the evaluation results while maintaining resolution. By using objective weights determined by game theory to perform a weighted summation of the correlation coefficients, a weighted correlation degree that comprehensively reflects the power quality level of each measuring point is obtained.

[0094] The final grey relational analysis results show that measuring point 3 achieved the highest weighted correlation degree of 0.766, indicating its strongest correlation with the ideal power quality state. Measuring point 5 achieved a correlation degree of 0.739, also demonstrating good power quality characteristics. Measuring point 1, with a correlation degree of only 0.658, ranked last among all measuring points, further confirming the severity of the power quality problem at this measuring point.

[0095] Step 5: The variance of the evaluation results of the fuzzy VIKOR method is 0.032, while the variance of the grey relational analysis is 0.018, indicating that the grey relational analysis has higher evaluation stability in this case.

[0096] Based on the principle of inverse variance-proportional weight allocation, grey relational analysis yielded a fusion weight of 0.887, while fuzzy VIKOR evaluation obtained a weight of 0.113. This weight allocation result fully reflects the essential characteristics of the traction power supply system as a grey system, verifying the applicability and advantages of the grey relational analysis method in this type of evaluation.

[0097] Step Six: Based on the power quality index grading values ​​defined by relevant national standards, the scoring ranges [0, 0.3319], [0.3319, 0.6308], [0.6308, 0.8923], and [0.8923, 1] were obtained following the steps above. The combined score calculation results show that measuring point 3, with the highest score of 0.993, ranks first and receives an "Excellent" rating, reflecting its excellent operating condition as a direct power supply device. Measuring point 5 scored 0.936, also reaching the "Excellent" level, reflecting the high reliability requirements of the signal system's power supply. Measuring point 4 scored 0.843, obtaining a "Good" level, indicating that its power quality is generally good but still has room for improvement.

[0098] According to the steps described in Example 2, the specific results of the power quality assessment are as follows:

[0099] Table 4. Power Quality Index Classification

[0100] measuring point Fuzzy VIKOR score Grey relational score Integrated score Evaluation results 3 0.516 0.766 0.993 excellent 5 0.522 0.739 0.936 excellent 4 0.433 0.746 0.843 good 2 0.500 0.556 0.485 medium 1 0.116 0.658 0.235 Needs improvement

[0101] Based on the traction substation measurement point layout diagram and the three-layer fusion evaluation results, measurement points 3 and 5 ranked first and second with fusion comprehensive scores of 0.993 and 0.936 respectively, both achieving an "excellent" rating. This result fully reflects the excellent operating status of the load-side equipment and signal power supply system. The outstanding performance of measurement point 3 reflects its good power quality stability when undertaking traction load power supply tasks, thanks to the effective role of the power quality monitoring device and dynamic compensation equipment configured on the load side. The excellent rating of measurement point 5 reflects that the reliability requirements of the signal system power supply have been effectively guaranteed. As a dedicated power supply for railway signal equipment, the stability of its power quality is directly related to train operation safety. The excellent evaluation results show that the signal power supply system is reasonably designed and operates stably. Measurement point 4 ranked third with a score of 0.843, obtaining a "good" rating. Although its overall performance is acceptable, its fuzzy VIKOR score of 0.433 is relatively low, indicating that there is still room for improvement in certain specific indicators.

[0102] The evaluation results highlight the performance of measuring points 1 and 2, which were rated "Needs Improvement" and "Medium" respectively, with evaluation scores of 0.235 and 0.485. As the input terminal of the 110kV high-voltage power supply, the root cause of power quality issues at measuring point 1 lies in the direct transmission of fluctuations in the operating status of the upstream power grid and power quality disturbances. This measuring point's fuzzy VIKOR score is only 0.116, far lower than other measuring points, indicating a significant deficiency in addressing voltage fluctuations and three-phase imbalances. The medium rating of measuring point 2 reflects the challenges of power quality control in medium-voltage distribution systems. As a power source for auxiliary equipment and some traction loads within the station, this measuring point needs to simultaneously ensure power supply reliability and power quality. However, its power quality control capabilities still need improvement when facing complex load characteristics and system disturbances.

[0103] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the invention can be implemented in other specific forms without departing from its spirit or essential characteristics. Therefore, the embodiments should be considered in all respects as exemplary and non-limiting, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of equivalents of the claims are intended to be included within the present invention. No reference numerals in the claims should be construed as limiting the scope of the claims.

Claims

1. A high-speed railway power quality assessment system based on three-layer fusion weighting, characterized in that, The system includes a game-theoretic dynamic weighting mechanism, fuzzy VIKOR evaluation theory, grey relational analysis, and a variance-based adaptive fusion mechanism; among which, The game theory dynamic weighting mechanism is the basic layer. By treating each power quality indicator as a rational game subject, it automatically identifies the key indicators with the strongest distinguishing ability in the high-speed railway operating environment and realizes the objective dynamic allocation of weights. The fuzzy VIKOR evaluation theory is the core layer. The triangular fuzzy number theory is used to model the uncertainty in the power quality measurement process of high-speed railway, and effectively handle the evaluation fuzziness caused by traction load fluctuations and changes in environmental factors. The grey relational analysis is a supplementary layer that deeply explores the correlation between each measuring point and the ideal operating state, and fully considers the grey system characteristics of the high-speed railway power supply system. The variance-based adaptive fusion mechanism enables collaborative decision-making among multiple methods, providing a basis for the assessment and systematic improvement of power quality in high-speed railways.

2. The high-speed railway power quality assessment system based on three-layer fusion weighting as described in claim 1, characterized in that, In the dynamic weighting mechanism of game theory, considering the dynamic characteristics and interrelationships of power quality indicators in the high-speed railway traction power supply system, m power quality indicators are regarded as rational game participants, and the payoff function of each indicator reflects its importance under the changing traction load environment: (1)； In the formula Let w be the revenue function value of the i-th index in the traction power supply system. i w is the weight of the i-th indicator. j Let σ be the weight of the j-th indicator. i ² Let ρ be the variance of the i-th index under different traction load conditions. ij The correlation coefficient between indicators in the traction power supply system, and the adjustment parameter α i =1、β i =0.5、γ i =0.1 controls the impact of discrimination gain, relevance competition, and weight penalty, respectively; The design of this revenue function takes into account the impact of rapid changes in the traction power of high-speed trains on the importance of various power quality indicators, and achieves automatic weight allocation that adapts to the dynamic characteristics of the traction power supply system through Nash equilibrium solution.

3. The high-speed railway power quality assessment system based on three-layer fusion weighting as described in claim 2, characterized in that, In the fuzzy VIKOR evaluation theory, considering the multiple uncertainties in the power quality measurement process of high-speed railway traction power supply systems, such as traction load fluctuations, dynamic contact between the overhead contact line and pantograph, and changes in environmental factors, the traditional VIKOR method needs to be improved by fuzzification to adapt to the complex operating environment of the traction power supply system. For the j-th power quality index at the i-th measurement point, a triangular fuzzy number is constructed: (2)； Among them, F̃ ij To account for the triangular fuzzy number of the traction power supply system uncertainty, x ij The standardized index value, σ j This represents the standard deviation of the index under different traction conditions. The construction of fuzzy numbers fully considers the impact of different operating conditions of high-speed trains—starting, braking, and constant-speed operation—on power quality. The group utility S is calculated using fuzzy distance. k and individual regrets R k The VIKOR comprehensive index of the k-th measurement point is obtained as follows: (3)； This formula balances the two dimensions of group utility and individual regret in the evaluation of traction power supply systems. It considers both the overall power quality level of each traction substation and the prominent problems of individual substations in specific indicators, where S * S — R represents the minimum and maximum group utility, respectively. * R — These represent the minimum and maximum values ​​of individual regrets, respectively. This balancing mechanism is particularly suitable for assessment needs in traction power supply systems that require both ensuring overall power supply quality and preventing individual components from becoming limiting factors.

4. The high-speed railway power quality assessment system based on three-layer fusion weighting according to claim 3, characterized in that, In grey relational analysis, the high-speed railway traction power supply system, as a typical grey system, exhibits characteristics of being neither completely known nor completely unknown due to the influence of various factors such as train schedules, traction load distribution, and equipment status. The variation patterns of power quality in the traction power supply system are often difficult to describe precisely using deterministic mathematical models. Grey relational analysis can effectively handle this characteristic of systems with incompletely determined information. Grey relational analysis is used to uncover the deep correlation between each traction substation and the ideal power supply state. The formula for calculating the correlation coefficient of the j-th power quality index at the k-th measuring point is: (4)； In the formula, ξ k (j) is the correlation coefficient between the traction substation and the ideal power supply state, Δ k (j) represents the absolute difference between the j-th index of the k-th traction substation and the ideal value, Δ min and Δ max These are the minimum and maximum values ​​of all absolute differences, respectively; The weighted correlation degree of the k-th measurement point is obtained by weighted summation based on the weights determined by game theory: (5)； In the formula, The weighted correlation degree of the k-th measurement point. Let be the weight of the j-th indicator.

5. The high-speed railway power quality assessment system based on three-layer fusion weighting according to claim 4, characterized in that, In the variance-based adaptive weight fusion mechanism, to fully leverage the different advantages of fuzzy VIKOR and grey relational analysis in power quality assessment of traction power supply systems and improve the reliability of results, a variance-based adaptive weight fusion mechanism is designed. This mechanism considers the diversity and complexity of the operating conditions of the traction power supply system, and judges its reliability by evaluating the dispersion of the results of each method under different traction load conditions. The larger the variance, the stronger the discrimination ability of the method under the current traction power supply system state. The final comprehensive score of the k-th measurement point is: (6)； In the formula, This represents the final comprehensive score for the k-th measurement point. The fuzzy VIKOR score of the kth measurement point The weights are the weights of the fuzzy VIKOR score for the k-th measurement point, obtained through a variance-based adaptive weight fusion mechanism. The grey relational analysis score of the k-th measurement point The weights of the grey relational analysis score of the k-th measurement point are obtained based on the variance-adaptive weight fusion mechanism. This adaptive mechanism ensures that the fusion evaluation results can reflect the power quality level of each traction substation under different traction load distributions and operating conditions.

6. A high-speed railway power quality assessment system based on three-layer fusion weighting, characterized in that, Includes the following steps: Step 1: Data preprocessing. Load power quality monitoring data from each measuring point, including seven core indicators: voltage fluctuation, three-phase voltage imbalance, voltage deviation, total harmonic distortion rate, odd voltage harmonics, even voltage harmonics, and frequency deviation. Clean the raw data, remove outliers and fill in missing data, and then perform normalization and standardization to map all indicators to the 0-1 range, laying the data foundation for subsequent evaluation and analysis. Step 2: Game theory dynamic weight allocation. The seven power quality indicators are regarded as rational game participants. A payoff function is constructed that considers the indicator's distinguishing ability, correlation competition and weight penalty. The objective weight of each indicator is solved by Nash equilibrium optimization algorithm to automatically identify the key indicator with the most distinguishing ability in the current data environment. Step 3: Fuzzy VIKOR evaluation. Construct triangular fuzzy numbers to handle uncertainties in the evaluation process, convert standardized index values ​​into fuzzy number form to account for measurement errors and environmental interference; determine positive and negative ideal solutions as reference benchmarks, calculate the group utility value and individual regret value of each measurement point, and finally generate a VIKOR comprehensive index that balances multi-objective conflicts to obtain the evaluation results based on fuzzy theory. Step 4: Grey relational analysis. Construct an ideal power quality state as a reference sequence, calculate the grey relational coefficient between each measuring point and the reference sequence on each index; use the weights determined by game theory to perform a weighted summation of the relational coefficients to obtain the comprehensive relational degree, and quantify the similarity and correlation between each measuring point and the ideal state from the perspective of grey system theory. Step 5: Adaptive fusion evaluation. Analyze the evaluation stability of the two methods, fuzzy VIKOR and grey relational analysis, and evaluate the reliability of each method through the variance index. Based on the reliability, dynamically allocate fusion weights, and combine the evaluation results of the two methods in a weighted manner to generate the final comprehensive evaluation score and ranking, giving full play to the advantages of each method and avoiding the limitations of a single method. Step 6: Results analysis. Based on the integrated evaluation scores, each measuring point is ranked and classified into levels, and the results are compared and verified with the national power quality standards. Based on the evaluation results, the main problems and weaknesses of each measuring point are identified, and the power quality evaluation is finally completed.

Images