A wind-solar-storage system for use in urban railway passenger stations

Abstract

This invention relates to the field of energy technology for railway passenger stations, and discloses a wind-solar-storage system applied to urban railway passenger stations. The system includes an environmental monitoring module that collects real-time data on wind speed, solar intensity, and load power demand; a renewable energy input balance assessment module that divides monitoring periods and calculates the wind energy input balance, solar energy input balance, and load demand fluctuation within each period; a dynamic load assessment module that generates dynamic load assessment values ​​based on the aforementioned balance and fluctuation; a multi-source coordinated control module that adjusts the blade pitch angle of the wind turbine generator, the tilt angle of the photovoltaic array, and the charging and discharging power of the energy storage unit according to the assessment values; and an operation strategy optimization module that optimizes operating parameters based on historical data. Through the coordinated operation of these modules, the system improves its operational stability and adaptability in urban railway passenger stations, reduces energy waste, and contributes to the green and low-carbon development of the stations.

Description

Technical Field

[0001] This invention relates to the field of railway passenger station energy technology, specifically a wind-solar-storage system applied to urban railway passenger stations. Background Technology

[0002] With the rapid expansion of urban rail networks, passenger stations, as transportation hubs with high passenger flow, are experiencing a continuous increase in total energy consumption, placing higher demands on the stability and sustainability of energy supply. Traditional urban rail passenger stations rely heavily on the municipal power grid for power supply, facing not only high grid load pressure and transmission losses, but also a lingering dependence on fossil fuels, which clashes with current green and low-carbon development concepts.

[0003] To alleviate this contradiction, some urban railway passenger stations have begun to explore the introduction of new energy sources such as wind and solar power, combining them with energy storage equipment to form wind-solar-storage systems. However, existing wind-solar-storage systems have revealed many problems in practical applications: wind and solar power are significantly intermittent and volatile, with their output power fluctuating dramatically with changes in the natural environment. Meanwhile, the electricity load of urban railway passenger stations also varies over time, leading to frequent imbalances between new energy supply and load demand, affecting the stability of the wind-solar-storage system. Furthermore, existing systems lack accurate assessment of the dynamic changes in new energy input characteristics and load demand, making it difficult to coordinate and control wind turbines, photovoltaic arrays, and energy storage units according to actual operating conditions, often resulting in energy waste or insufficient supply. In addition, most wind-solar-storage systems have not established effective operational strategy optimization mechanisms, making it impossible to continuously adjust equipment operating parameters based on historical data. This makes it difficult to adapt to changes in the environment and load during long-term use, limiting the promotion and application of wind-solar-storage systems in urban railway passenger station scenarios. Summary of the Invention

[0004] The purpose of this invention is to provide a wind-solar-storage system for use in urban railway passenger stations, in order to solve the problems mentioned in the background art.

[0005] To achieve the above objectives, the present invention provides a wind-solar-storage system for use in urban railway passenger stations, comprising:

[0006] The environmental monitoring module is used to collect real-time data on wind speed, light intensity, and load power demand in the area where the city railway passenger station is located.

[0007] The new energy input balance assessment module is used to receive the wind speed data, the light intensity data and the load power demand data, divide the monitoring period, and calculate the wind energy input balance, light energy input balance and load demand fluctuation within the preset standard range in each monitoring period.

[0008] The dynamic load assessment module is used to generate dynamic load assessment values ​​for each monitoring period based on the wind energy input balance, solar energy input balance, and load demand fluctuation.

[0009] The multi-source coordinated control module is used to adjust the blade pitch angle of the wind turbine generator, the tilt angle of the photovoltaic array, and the charging and discharging power of the energy storage unit based on the dynamic load assessment values ​​in each monitoring cycle.

[0010] The operation strategy optimization module is used to optimize the operation parameters based on historical operation data after the preset time window ends.

[0011] Furthermore, the wind energy input uniformity is obtained through the following steps:

[0012] Extract the maximum and minimum values ​​of the wind speed data within each monitoring period, and calculate the wind speed fluctuation difference;

[0013] Obtain a preset standard wind speed fluctuation range, and combine the wind speed fluctuation difference with the standard wind speed fluctuation range to generate the wind energy input balance.

[0014] Furthermore, the optical energy input equalization is obtained through the following steps:

[0015] Extract the highest and lowest values ​​of the light intensity data within each monitoring period, and calculate the light intensity fluctuation difference;

[0016] A preset standard light intensity fluctuation range is obtained, and the light energy input balance is generated by combining the light intensity fluctuation difference and the standard light intensity fluctuation range.

[0017] Furthermore, the load demand fluctuation is obtained through the following steps:

[0018] Extract the peak and valley values ​​of the load power demand data within each monitoring period, and calculate the load demand fluctuation difference;

[0019] Obtain a preset standard load fluctuation range, and combine the load demand fluctuation difference with the standard load fluctuation range to generate the load demand fluctuation degree.

[0020] Furthermore, dynamic load assessment values ​​are generated for each monitoring period, specifically:

[0021] The wind energy input balance, solar energy input balance, and load demand fluctuation within each monitoring period are weighted and calculated to output the dynamic load assessment value of the wind-solar-storage system.

[0022] Furthermore, adjusting the blade pitch angle of the wind turbine generator, the tilt angle of the photovoltaic array, and the charging and discharging power of the energy storage unit based on the dynamic load assessment values ​​within each monitoring period specifically includes:

[0023] Compare the dynamic load assessment values ​​within each monitoring period with the pre-stored dynamic load assessment threshold range;

[0024] If the dynamic load assessment value of the wind-solar-storage system is higher than the upper limit of the pre-stored dynamic load assessment threshold range during the current monitoring period, then the blade pitch angle of the wind turbine generator is increased, the tilt angle of the photovoltaic array is decreased, and the discharge power of the energy storage unit is increased.

[0025] If the dynamic load assessment value of the wind-solar-storage system is lower than the lower limit of the pre-stored dynamic load assessment threshold range during the current monitoring period, the blade pitch angle of the wind turbine generator is reduced, the tilt angle of the photovoltaic array is increased, and the charging power of the energy storage unit is improved.

[0026] Furthermore, it also includes a safety linkage module for performing the following operations:

[0027] When the real-time wind speed collected by the environmental monitoring module exceeds the preset wind speed safety threshold, the blade pitch angle of the wind turbine generator set is locked.

[0028] When the real-time light intensity collected by the environmental monitoring module is lower than the preset light intensity safety threshold, the photovoltaic array is switched to the backup power supply mode.

[0029] When the instantaneous value of the load power demand collected by the environmental monitoring module exceeds the preset load safety threshold, the peak compensation function of the energy storage unit is activated.

[0030] Furthermore, the specific process of the optimization performed by the operation strategy optimization module includes:

[0031] Extract the temporal characteristics of wind speed data, light intensity data, and load power demand data from the historical operating data;

[0032] A future operational trend prediction model is established based on the aforementioned time-series characteristics;

[0033] The reference values ​​for the blade pitch angle of the wind turbine generator, the tilt angle of the photovoltaic array, and the charging and discharging power of the energy storage unit are corrected based on the future operation trend prediction model.

[0034] Furthermore, the operation strategy optimization module also includes:

[0035] A historical operation feature database is constructed to store the blade pitch angle adjustment records of the wind turbine generator, the tilt angle adjustment records of the photovoltaic array, and the charging and discharging power adjustment records of the energy storage unit;

[0036] When the cumulative number of times the blade pitch angle adjustment record, tilt angle adjustment record, and charge / discharge power adjustment record reaches the preset adjustment count threshold, a parameter calibration command is triggered.

[0037] Furthermore, when the operation strategy optimization module responds to the parameter calibration command, it performs the following:

[0038] Extract the environmental monitoring data difference values ​​of adjacent monitoring periods from the historical operation feature library;

[0039] The pre-stored dynamic load assessment threshold range is adaptively corrected based on the differences in the environmental monitoring data.

[0040] Compared with the prior art, the beneficial effects of the present invention are:

[0041] The environmental monitoring module collects real-time data on wind speed, light intensity, and load power demand, providing comprehensive and timely basic information for subsequent assessment and control.

[0042] The new energy input balance assessment module divides the monitoring period based on the collected data and calculates the wind energy input balance, solar energy input balance and load demand fluctuation in each period. It can accurately capture the dynamic characteristics of new energy supply and load demand and clearly reflect the matching status of the two in different time periods.

[0043] The dynamic load assessment module combines the above-mentioned balance and volatility to generate dynamic load assessment values, providing a scientific basis for the coordinated control of wind, solar and energy storage systems, and making the control measures more in line with actual operating conditions.

[0044] Based on dynamic load assessment values, the multi-source coordinated control module adjusts the blade pitch angle of the wind turbine generator, the tilt angle of the photovoltaic array, and the charging and discharging power of the energy storage unit in a targeted manner. This enables the coordinated operation of new energy power generation equipment and energy storage equipment, effectively addressing the intermittency of new energy supply and the volatility of load demand, and reducing the imbalance between energy supply and demand.

[0045] After a preset time window expires, the operation strategy optimization module optimizes the operating parameters of the wind-solar-storage system based on historical operating data. This allows the equipment to gradually adapt to changes in the environment and load during long-term use, improving the adaptability of the wind-solar-storage system under different time periods and operating conditions. Overall, through the synergistic effect of its various modules, the equipment enhances the stability and adaptability of the wind-solar-storage system in urban railway passenger station scenarios, reduces energy waste, and allows new energy sources to play a more effective role in the energy supply of passenger stations, contributing to the green and low-carbon development of urban railway passenger stations. Attached Figure Description

[0046] Figure 1 This is a timing diagram of the wind-solar-storage system applied to urban railway passenger stations as described in this invention;

[0047] Figure 2 A flowchart for calculating the wind energy input balance.

[0048] Figure 3 A flowchart for adjusting multi-source coordinated control;

[0049] Figure 4 The flowchart for the execution of the safety linkage module;

[0050] Figure 5 A flowchart for the execution of the strategy optimization module. Detailed Implementation

[0051] 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 skilled in the art without creative effort are within the scope of protection of the present invention.

[0052] Please see Figure 1 This invention provides a wind-solar-storage system for use in urban railway passenger stations. The wind-solar-storage system includes: an environmental monitoring module, a new energy input balance assessment module, a dynamic load assessment module, a multi-source coordination control module, and an operation strategy optimization module.

[0053] The environmental monitoring module collects real-time data on wind speed, solar intensity, and load power demand, providing input parameters for the wind-solar-storage system. The renewable energy input balance assessment module divides the monitoring period and calculates the wind energy input balance, solar energy input balance, and load demand fluctuation. The dynamic load assessment module generates dynamic load assessment values ​​based on these balance and fluctuation values. The multi-source coordinated control module adjusts the wind turbine blade pitch angle, photovoltaic array tilt angle, and energy storage unit charging and discharging power according to the dynamic load assessment values. After the time window ends, the operation strategy optimization module optimizes operating parameters based on historical operating data to improve overall operating efficiency.

[0054] The operating parameters refer to the blade pitch angle reference value, the photovoltaic array tilt angle reference value, the energy storage unit charge and discharge power reference value, and the pre-stored dynamic load evaluation threshold.

[0055] Example 1: The monitoring cycle is preset according to the operational characteristics of urban railway passenger stations, and data acquisition units are usually divided into fixed time intervals. (See the flowchart for wind energy input balance calculation.) Figure 2The method involves automatically extracting the maximum and minimum wind speed data within each monitoring cycle and calculating the difference between them to obtain the wind speed fluctuation difference. This difference reflects the fluctuation range of wind energy input within the current cycle. A preset standard wind speed fluctuation range serves as the evaluation benchmark, based on historical meteorological data and the operating parameters of the wind turbine generators. The wind speed fluctuation difference is compared and analyzed with the standard wind speed fluctuation range: if the difference is within the standard wind speed fluctuation range, it indicates that the wind energy input is relatively stable; if it exceeds the upper limit of the standard wind speed fluctuation range, it indicates that the wind energy fluctuation is drastic; if it is below the lower limit of the standard wind speed fluctuation range, it indicates that the wind energy input changes relatively little. Specifically, when the difference is within the standard wind speed fluctuation range, the difference is normalized and quantified based on the standard wind speed fluctuation range to generate a wind energy input balance degree. The numerical range of the wind energy input balance degree is set between 0 and 1, with a higher value indicating better balance. Using the preset lower limit (Smin) and upper limit (Smax) of the standard wind speed fluctuation range as benchmarks, calculate the width of the standard wind speed fluctuation range (i.e., Smax - Smin); when the wind speed fluctuation difference (Δv) within the current monitoring period is within the interval [Smin, Smax], calculate the wind energy input balance using the following formula:

[0056]

[0057] Specifically, if Δv equals Smax (i.e., maximum fluctuation), the normalized result is 0; if Δv equals Smin (i.e., minimum fluctuation), the normalized result is 1. When Δv falls between these two values, the result is linearly inversely mapped to the 0-1 interval. This converts the wind speed fluctuation difference into a quantitative indicator reflecting the balance of wind energy input. A higher value indicates smaller wind speed fluctuations and more stable wind energy input within that period. The standard wind speed fluctuation range used in this process includes the initial range based on historical meteorological data and wind turbine parameters, as well as the latest range adjusted through a calibration procedure.

[0058] When the difference is outside the standard wind speed fluctuation range, wind speed fluctuation difference calibration is performed without normalization. Specifically, when the wind speed fluctuation difference is outside the standard wind speed fluctuation range, the operating status of the wind speed sensor and the data acquisition process are first checked for any abnormalities. If the difference is abnormal due to a momentary sensor malfunction or data transmission problem, the forward filling algorithm is activated, using the valid wind speed data from the previous period to replace the abnormal data, and the wind speed fluctuation difference is recalculated. If the sensor and acquisition process are normal, the latest fluctuation data is incorporated into the historical dataset based on the wind speed data recently collected by the environmental monitoring module, using the moving average algorithm, and the standard wind speed fluctuation range is recalculated and adjusted. Then, the wind energy input balance is regenerated based on the adjusted standard wind speed fluctuation range. When the wind energy input balance jumps beyond a set amplitude in adjacent monitoring periods, the wind speed sensor calibration check is automatically triggered.

[0059] The light intensity data collected by the environmental monitoring module is divided into the same monitoring period. Within each period, the highest and lowest values ​​of the light intensity data are identified, and the absolute difference between the two is used to obtain the light intensity fluctuation difference. The preset standard light intensity fluctuation range is derived from local meteorological characteristics and photovoltaic equipment technical parameters. The light intensity fluctuation difference is compared and analyzed with the standard light intensity fluctuation range: when the difference is within the standard light intensity fluctuation range, the light energy input state is judged to be stable; when the difference is higher than the upper limit of the standard light intensity fluctuation range, it indicates that the light intensity changes drastically; when the difference is lower than the lower limit of the standard light intensity fluctuation range, it indicates that the light intensity changes slightly. Specifically, when the difference is within the standard light intensity fluctuation range, the difference is normalized and quantified based on the standard light intensity fluctuation range to generate a light energy input balance degree. The light energy input balance degree value is also in the range of 0 to 1, and the value directly reflects the stability level of photovoltaic power generation input. When the difference is not within the standard light intensity fluctuation range, light intensity fluctuation difference calibration is performed without normalization. Specifically: When the light intensity fluctuation difference is outside the standard light intensity fluctuation range, first check whether the light monitoring sensor is obstructed or malfunctioning. If the data anomaly is caused by obstruction, initiate a linear interpolation compensation method to correct the light intensity data and recalculate the fluctuation difference. If the sensor is operating normally, combine the light intensity data recently collected by the environmental monitoring module, adjust the standard light intensity fluctuation range using a moving average algorithm, and then recalculate the light energy input balance based on the adjusted standard range. When the light energy input balance remains below the warning value, the wind-solar-storage system initiates the photovoltaic panel cleanliness detection program.

[0060] The load power demand data collected in real time by the environmental monitoring module is divided into monitoring cycles synchronized with the wind and solar power assessment. Within each monitoring cycle, the peak and trough values ​​of the load power demand data are extracted, and the absolute difference between the two is used to obtain the load demand fluctuation difference. The preset standard load fluctuation range is set with reference to the historical power consumption curves of the passenger station and the equipment's carrying capacity. The load demand fluctuation difference is compared and analyzed with the standard load fluctuation range: if the difference is within the standard load fluctuation range, it indicates that the load demand change is within a normal fluctuation range; if the difference exceeds the upper limit of the standard load fluctuation range, it indicates that the load has experienced an abnormal peak; if the difference is below the lower limit of the standard load fluctuation range, it indicates that the load demand is abnormally stable. Specifically, when the difference is within the standard load fluctuation range, the difference is normalized and quantified based on the standard load fluctuation range to generate a load demand fluctuation degree. The load demand fluctuation degree value is also within the range of 0 to 1, with a higher value indicating stronger load volatility. When the difference is not within the standard load fluctuation range, load fluctuation difference calibration is performed without normalization. Specifically, when the load demand fluctuation difference is outside the standard load fluctuation range, the system checks for communication interruptions in the load power demand data acquisition channel. If a communication anomaly is found, the system retrieves the load power demand data from the nearest neighboring period for supplementation and correction, and recalculates the fluctuation difference. If the acquisition channel is normal, the system updates the standard load fluctuation range using a moving average algorithm based on the recently collected load power demand data from the environmental monitoring module, and regenerates the load demand fluctuation according to the updated standard range. When the load demand fluctuation abnormally increases, a list of recommended equipment inspections is generated.

[0061] The monitoring cycle starts at a uniformly triggered time, ensuring complete overlap in the collection periods for wind speed, light intensity, and load power demand data. The environmental monitoring module includes a data preprocessing stage, which incorporates outlier filtering: when the wind speed sensor collects outlier data that momentarily exceeds the safety threshold, a forward-fill algorithm is automatically activated, replacing the outlier with valid values ​​from the previous period; when the light intensity sensor records zero values ​​due to obstruction, linear interpolation compensation is initiated; and when communication is interrupted during load power demand data collection, the nearest neighbor period's data is used for supplementation.

[0062] The fluctuation difference is calculated using a rolling window algorithm, which automatically updates the extreme value records at the beginning of each new cycle to avoid the accumulation of historical data. Specifically, the extreme value records include: the maximum and minimum wind speed values ​​extracted during the calculation of wind energy input balance; the highest and lowest light intensity values ​​extracted during the calculation of solar energy input balance; and the peak and valley values ​​of load electricity demand data extracted during the calculation of load demand fluctuation. These extreme value data, obtained through data acquisition and preprocessing, are recorded by the system to form "extreme value records," providing a data basis for automatically updating the extreme values ​​when using the rolling window algorithm to calculate the fluctuation difference.

[0063] The initial standard wind speed fluctuation range is set based on the annual average wind speed distribution at the installation location, the standard light intensity fluctuation range is referenced from the local annual solar radiation variation curve, and the standard load fluctuation range is generated based on the passenger station's peak and valley electricity consumption data for the past three years. The environmental monitoring module automatically analyzes the latest collected environmental data monthly. When the proportion of data whose actual fluctuation range is outside the initial standard range exceeds a set percentage within three consecutive sampling periods, the standard range calibration procedure is initiated. Specifically, taking wind speed data as an example, the initial standard wind speed fluctuation range is set based on the annual average wind speed distribution at the installation location (e.g., if the annual average wind speed is 3.5 m / s, the standard wind speed fluctuation range is set to 2.0-5.0 m / s). If, in three consecutive sampling periods (each period is 10 minutes, and 30 wind speed data points are collected per period), in the first period, 18 data points have actual fluctuation ranges (e.g., wind speed fluctuation differences of 6.2 m / s, 4.8 m / s, etc.) that are not within the initial standard range of 2.0-5.0 m / s, accounting for 60%; in the second period, 21 data points exceed this range, accounting for 70%; and in the third period, 15 data points exceed this range, accounting for 50%. Here, the "set ratio" can be set to 55%, because the data proportions of the first two periods exceed 55% in all three periods, which meets the conditions for starting the standard range calibration procedure. The calibration process will recalculate the standard range by incorporating the latest fluctuation data from these three periods through a moving average algorithm. Standard range calibration procedure: The calibration process uses a moving average algorithm to recalculate the latest fluctuation data by incorporating it into the historical dataset to obtain the latest standard range. Then, the corresponding equilibrium and volatility are quantitatively calculated using the latest standard range.

[0064] For wind energy input balance, when the wind speed fluctuation difference is in the middle of the standard range, the balance is linearly negatively correlated with the difference; when the difference approaches or exceeds the range boundary, the rate of change of balance gradually decreases. Specifically, when the wind speed fluctuation difference is within the preset standard wind speed fluctuation range and does not approach the middle interval of the upper and lower boundaries of the range, the change in the magnitude of the wind speed fluctuation difference will directly lead to an inverse linear change in wind energy input balance. The larger the wind speed fluctuation difference, the greater the fluctuation amplitude of wind energy input within the monitoring period, the worse the stability of wind energy input, and the lower the corresponding wind energy input balance; the smaller the wind speed fluctuation difference, the smaller the fluctuation amplitude of wind energy input, the better the stability, and the higher the wind energy input balance. The conversion curve for solar energy input balance sets a solar intensity compensation factor, which automatically widens the standard range boundary when the sunshine duration is lower than the seasonal average. Here, the conversion curve is a quantitative mapping relationship between the solar intensity fluctuation difference and the solar energy input balance; the solar intensity compensation factor is a seasonal conditional parameter that triggers the widening of the standard range. The normalization of load demand fluctuations incorporates a time-weighted coefficient, allowing a wider range of fluctuations during weekday peak hours compared to nighttime hours. All normalization results are rounded to two decimal places.

[0065] The data processing mechanism for the collaborative processing of the new energy input balance assessment module and the dynamic load assessment module is divided into a basic layer, an intermediate layer, and an application layer. This layer standardizes the processing and output logic of data related to wind energy input balance, solar energy input balance, and load demand fluctuation. The basic layer contains the original fluctuation difference and standard range boundary values; the intermediate layer contains the balance / fluctuation values; and the application layer adds trend identifiers. When the dynamic load assessment values ​​change in the same direction for three consecutive monitoring periods, exceeding the pre-stored dynamic load assessment threshold range, an "upward trend" or "downward trend" status is automatically marked. Data transmission uses a lightweight encryption protocol, with each data packet containing a timestamp, device number, and checksum to prevent tampering during transmission. The dynamic load assessment threshold range is preset based on historical assessment value fluctuation characteristics, equipment response sensitivity, and station energy regulation requirements. It represents the cumulative amplitude threshold value for the same-direction change of wind energy input balance, solar energy input balance, or load demand fluctuation (all quantized values ​​in the 0-1 range) over three consecutive periods. All assessment data is stored with environmental tags, including auxiliary parameters such as temperature and humidity, for in-depth analysis.

[0066] During implementation, wind speed monitoring points were placed in an open area on the passenger station roof at a set height above the ground, avoiding the building's turbulence zone; light monitoring sensors were installed on an unobstructed platform and their angles were calibrated regularly; load monitoring utilized a cluster of smart meters, with sampling frequencies synchronized with the wind and solar monitoring equipment. The environmental monitoring module also included a data quality dashboard, displaying the signal strength and confidence level indicators of each monitoring point in real time. When the confidence level of a certain data point fell below a threshold, the system automatically switched to a backup monitoring point and equipment.

[0067] Raw wind speed data is first stored in a circular buffer, and an extreme value extraction thread runs independently to avoid blocking the main process. Light intensity processing employs a parallel computing architecture, with data acquisition and extreme value identification occurring simultaneously. A dedicated transmission channel is established for load demand data, and a priority scheduling algorithm ensures the real-time performance of peak power consumption data. All calculation processes are equipped with timeout interrupt protection; if a single cycle times out, a simplified algorithm is activated and an exception log is generated.

[0068] The wind-solar-storage system of this invention also reserves a meteorological data interface and an evaluation result output port. The meteorological data interface can access extreme weather warning information issued by the local meteorological station. When a strong wind or rainstorm warning is received, the standard range setting for wind speed fluctuation is temporarily relaxed. The normalized parameter table adopts dual backup storage, including a main parameter table and a backup parameter table. When the main parameter table is abnormal, it automatically switches to the backup parameter table to continue operation. The evaluation result output port is equipped with a data shaping filter to eliminate the jitter of evaluation values ​​caused by acquisition interference.

[0069] When wind speed data remains unchanged for multiple consecutive periods, the sensor is automatically flagged as faulty; when solar irradiance data deviates continuously from astronomical calculations, a calibration alert is triggered; and when load power demand data shows regular abrupt changes, line inspection suggestions are generated. This invention's wind-solar-storage system also includes a central monitoring center, where all diagnostic information is integrated in real time to form an equipment health status map.

[0070] Example 2: The dynamic load assessment module includes a weight redistribution program, a configurable parameter table, and a historical weight record library. Wind energy input balance, solar energy input balance, and load demand fluctuation are used as three assessment indicators. Weight coefficients are assigned to these three indicators, and the values ​​are stored in the configurable parameter table. Initially, the weight coefficients for wind energy input balance, solar energy input balance, and load demand fluctuation are set to 0.35 and 0.30, respectively, based on the energy characteristics of a typical urban railway passenger station. The weight coefficients can be dynamically adjusted in the background. When a certain indicator's data deviates abnormally from its historical average over multiple consecutive monitoring periods, the weight allocation is automatically corrected according to predetermined rules.

[0071] At the start of each monitoring cycle, the values ​​of three evaluation indicators are synchronously transmitted to the buffer in the dynamic load assessment module. The calculation unit first performs data type validation to exclude non-numerical erroneous data; then, it performs unit standardization processing to convert the input values ​​into standard floating-point format; finally, it calculates the evaluation value using a multiplicative summation method: wind energy input balance is multiplied by the corresponding weighting coefficient, solar energy input balance is multiplied by the corresponding weighting coefficient, and load demand fluctuation is multiplied by the corresponding weighting coefficient. The three sets of product results are added together, and the final evaluation value is output. The calculation result is retained to two decimal places, and the value range is controlled between 0.00 and 1.00.

[0072] The process of automatically correcting weight allocation according to predetermined rules is as follows: The historical weight record database stores the weight configuration records for the most recent thirty days. When the fluctuation range of a certain evaluation indicator exceeds a threshold for ten consecutive monitoring periods, the weight redistribution procedure is triggered. The redistribution process is based on a moving average algorithm: the coefficient of variation of the evaluation indicator over the most recent one hundred monitoring periods is calculated, and the weight coefficient ratio is adjusted in reverse according to the magnitude of the coefficient of variation. The higher the coefficient of variation, the greater the reduction in the weight coefficient of that indicator, with the upper limit of the reduction set at the original weight coefficient. At the same time, the weight coefficients of the other two indicators are increased in equal proportions according to their remaining values, keeping the total weight sum a constant value, i.e., a constant value of 1.

[0073] Each time the environmental monitoring module generates a set of data on wind speed, solar irradiance, and load power demand, it immediately adds a millisecond-level precise timestamp. Before performing calculations, the dynamic load assessment module verifies the consistency of the timestamps for the three assessment indicators; data sets with time deviations exceeding a set number of milliseconds are marked as invalid. The wind-solar-storage system also has a data waiting timeout mechanism: when data for a certain indicator arrives late, it waits for a maximum preset time; if the timeout occurs, the corresponding data from the previous cycle is used to complete the data. Data completion operations are recorded in the operation log; a set number of consecutive completion events will trigger a data channel maintenance alarm.

[0074] Before outputting the calculation results, a validity check is performed. The check rules include three aspects: checking whether the evaluated value is within the theoretical range; comparing the jump amplitude between the current value and the previous period value to see if it exceeds a reasonable threshold; and verifying whether the mathematical relationship between the input data and the output result conforms to the calculation logic. If any check fails, the system automatically switches to a backup calculation strategy: using the median of the three indicators instead of the weighted average, and adding an anomaly status indicator to the output data packet. The backup calculation results are stored in parallel with the main calculation results for later analysis and comparison.

[0075] The dynamic load assessment module comprises a main control layer, an auxiliary analysis layer, a data archiving layer, and an assessment value database. The main control layer receives raw assessment values ​​and generates equipment adjustment commands. The auxiliary analysis layer performs waveform analysis on the assessment value sequence to identify periodic fluctuation patterns. The data archiving layer stores complete historical assessment records, forming a time-series database. The assessment value database uses a partitioned storage structure, with data partitions established according to the monitoring period. Each partition stores the raw input data, weight configuration parameters, calculation process records, and final output values. The assessment value database supports fast retrieval by time range, with the smallest query unit being a complete monitoring period.

[0076] The dynamic load assessment module of this invention also includes an abnormal data processing mechanism, a backend management system, a main computing unit, a downward connection layer, an upward support layer, a horizontal collaboration layer, and a backup computing unit. The abnormal data processing mechanism includes multi-level responses. The first-level response addresses missing data for a single item: when wind energy input balance data is missing, it defaults to the average of the previous three periods; when solar energy input balance data is missing, it is filled with the historical average of the same day and time; when load demand fluctuation data is missing, it is estimated based on the passenger station's operating schedule. The second-level response addresses calculation errors: if wind energy input balance is detected to exceed 1.0, it is automatically truncated to 0.99; if solar energy input balance is negative, it is reset to zero; when load demand fluctuation abnormally increases, a data review process is initiated. The third-level response addresses computational interruption events: when insufficient hardware resources cause computation timeouts, the current period's computation is skipped, and the output value of the previous period is directly used.

[0077] The backend management system provides a visual panel for manual adjustment of various weight coefficients; it also features a seasonal mode switching function with preset summer / winter differentiated weight templates; and an open API interface allows external energy management systems to push weight adjustment commands. All parameter modifications require dual verification: after operator authorization, a dynamic security code must be entered for the changes to take effect. Parameter modification records are synchronized to the audit database in real time, including key fields such as operator information, modification time, and the change in values ​​before and after the modification.

[0078] The main processing unit uses a multi-core processor to process data in parallel, while the backup processing unit continuously receives the same input source. After outputting the result each cycle, the main processing unit sends a verification command to the backup processing unit. The backup processing unit executes the same calculation process; if the difference in the output result exceeds the set fault tolerance range, a collaborative verification procedure is initiated: the two sets of processing units exchange intermediate process data and compare the numerical consistency of each stage of multiplication, accumulation, and addition. The final output result is the value from the main processing unit, but the discrepancy event is recorded in the device operation status report.

[0079] Each data packet output by the dynamic load assessment module includes fields such as header identifier, time segment code, original values ​​of the three metrics, current weight configuration, calculation result value, calculation time, and data quality identifier. The data packets are transmitted to the multi-source coordination control module via an industrial bus, with dedicated bandwidth reserved for this purpose. The receiving end features dual buffers: the current period's data packet is stored in the working buffer, while the previous period's data packet remains available in the standby buffer. This design ensures that even with current data packet processing delays, there is still valid data available for device adjustment.

[0080] For high-frequency monitoring scenarios, when the monitoring period is less than a set time threshold, a simplified calculation strategy is automatically activated: weighting coefficients are rounded to a preset precision, floating-point processing is omitted during the calculation process, and the result is retained to one decimal place. When the simplified calculation mode is activated, the performance mode indicator is displayed in the status panel of the visualization panel. When the monitoring period returns to the normal range, it automatically switches back to the standard calculation mode. The computing unit temperature monitoring module continuously samples the processor temperature, and when it exceeds the warning temperature, it actively reduces the computing frequency and activates the heat dissipation enhancement program.

[0081] Every Sunday morning, batch data processing is automatically performed: summarizing the dynamic evaluation value sequence over seven days and generating a fluctuation trend chart; marking the time periods when the highest / lowest evaluation values ​​occur on a single day; and identifying the actual distribution ratio of the weights of the three indicators. Analysis reports are automatically pushed to the management terminal, allowing manual adjustments to the preset weight strategy for the following week. Monthly analysis includes correlation detection: calculating the correlation coefficient between the evaluation values ​​and equipment energy consumption to verify the effectiveness of the dynamic evaluation model. The dynamic evaluation model refers to the entire evaluation logic whereby the environmental monitoring module collects real-time data on wind speed, solar intensity, and load power demand at the city's railway passenger stations; the new energy input balance evaluation module divides the monitoring period and calculates the wind energy input balance, solar energy input balance, and load demand fluctuation within each period; and the dynamic load evaluation module weights these three indicators according to preset weights, ultimately outputting a dynamic load evaluation value in the range of 0.00 to 1.00. The analysis results are stored in a knowledge base to provide training samples for the weight self-learning algorithm. The weight self-learning algorithm initially presets the weight coefficients for wind energy input balance (0.35), solar energy input balance (0.35), and load demand fluctuation (0.30) based on the energy characteristics of the urban railway passenger station. When the fluctuation range of a certain indicator exceeds the threshold for several consecutive periods, its weight ratio is adjusted in reverse based on the dispersion coefficient of the indicator in the most recent hundred periods (the higher the dispersion coefficient, the greater the weight reduction). At the same time, the weights of the other two indicators are increased by an equal amount to keep the total weight of 1. The algorithm uses the analysis results of weekly and monthly dynamic evaluation values ​​as training samples to gradually adapt to the weight adjustment algorithm for environmental and load changes.

[0082] Each output data packet is augmented with a cyclic redundancy check (CRC) code, and the receiver performs CRC verification before parsing the data. Critical numerical fields undergo byte-reversal processing, preventing illegally intercepted data from being parsed correctly. The communication link maintains a heartbeat mechanism, sending handshake signals at set time intervals. If three consecutive handshakes fail, the connection is terminated, and the system automatically switches to a backup transmission channel. All communication failure events trigger real-time alarms, displaying channel repair instructions on the central monitoring center's screen.

[0083] The downward connection layer provides an open data subscription service, allowing the multi-source coordination control module to register and receive evaluation value data streams with specified precision. The upward support layer provides a historical data access interface, supporting queries based on multiple conditions such as time dimension, numerical threshold, and quality level. The horizontal collaboration layer sets up a cross-data exchange area, allowing the security linkage module to retrieve the latest evaluation values ​​for decision-making in emergency situations. All interface calls require credential verification; unauthorized wind, solar, and energy storage systems cannot access core evaluation data. The multi-source coordination control module refers to a set of functional modules that can obtain control basis by subscribing to dynamic load evaluation value data streams, thereby executing energy supply and demand balancing operations. Its core is the multi-source coordination control module and the adjustment mechanisms of the wind turbine generators, photovoltaic arrays, and energy storage units controlled by it. Its main function is to adjust the blade pitch angle of the wind turbine generators, the tilt angle of the photovoltaic arrays, and the charging and discharging power of the energy storage units based on the received dynamic load evaluation values ​​to match the new energy supply and load power demand of the urban railway passenger station.

[0084] Example 3: See Figure 3 The execution logic of the multi-source coordination control module is based on dynamic load assessment values. The comparison results with the pre-stored dynamic load assessment threshold range include the wind turbine generator control unit, photovoltaic array control unit, and energy storage unit control module. The pre-stored dynamic load assessment threshold range includes the upper limit value. and lower limit value These two parameters are initialized based on the device's historical operating data, among which... This indicates the maximum allowable dynamic load assessment value for a wind-solar-storage system. This represents the minimum permissible dynamic load assessment value for the wind, solar, and energy storage system. The comparison process is triggered at the end of each monitoring cycle, using the dynamic load assessment value calculated for the current cycle. and interval [ , Real-time comparison is performed to generate three types of adjustment instructions.

[0085] when At that time, it is determined that the current new energy input is fluctuating drastically or the load demand is too high. The wind turbine generator control unit receives a command to increase the blade pitch angle, and the pitch angle adjustment amount... The current wind speed v and the rated wind speed The ratio determines:

[0086]

[0087] In the formula: For wind turbine characteristic coefficients, The wind turbine is designed for its rated wind speed. This adjustment increases the blade angle of attack, reducing wind energy capture efficiency. The photovoltaic array control unit synchronously receives commands to reduce the tilt angle, and the tilt angle adjustment amount... Based on the current light intensity I and the standard light intensity I std The difference calculation gradually brings the photovoltaic panel closer to a horizontal position to reduce the area exposed to sunlight. The energy storage unit control module initiates a discharge power boosting program, with the discharge increment... The extent to which the assessed value exceeds the upper limit It is proportional to the output voltage waveform, which is adjusted by power electronic devices.

[0088] when At this time, it is determined that the new energy input is stable and the load demand is low. The wind turbine generator control unit executes an operation to reduce the blade pitch angle. The pitch angle adjustment uses the opposite algorithm to the above, so that the blades tend towards the optimal angle of attack to improve wind energy conversion efficiency. The photovoltaic array control unit calculates the increased tilt angle value and adjusts the orientation of the photovoltaic panels according to real-time solar altitude angle data to maximize the amount of light energy received. The energy storage unit switches to charging priority mode, increasing the charging power. The extent to which the assessed value is below the lower limit The relationship is linear, and the charging current curve is adjusted by a bidirectional converter.

[0089] The reverse algorithm is specifically explained as follows: when the dynamic load assessment value At that time, the wind turbine generator control unit performs an operation to reduce the blade pitch angle, and the pitch angle adjustment amount is... The algorithm used is the opposite of increasing the pitch angle, and the specific formula is as follows:

[0090]

[0091] in, This represents the blade pitch angle adjustment (a negative value indicates a decrease in the pitch angle). This is the characteristic coefficient of the wind turbine unit (consistent with the coefficient used in the algorithm for increasing the pitch angle, and set based on the inherent parameters such as the model and power of the wind turbine unit). The actual wind speed during the current monitoring period (collected by the environmental monitoring module); The wind turbine is designed with a rated wind speed (the optimal operating wind speed value set at the factory). This algorithm calculates the square of the difference between the rated wind speed and the current wind speed, multiplied by a negative wind turbine characteristic coefficient, so that the blade angle of attack decreases reasonably as the difference between the current wind speed and the rated wind speed decreases, gradually approaching the optimal wind energy capture angle, thereby improving wind energy conversion efficiency and adapting to operating conditions when renewable energy input is stable and load demand is low.

[0092] The blades are positioned at the optimal angle of attack to improve wind energy conversion efficiency. The photovoltaic array control unit calculates the increased tilt angle and adjusts the orientation of the photovoltaic panels based on real-time solar altitude angle data to maximize solar energy reception. The energy storage unit switches to charging priority mode, increasing charging power. The extent to which the assessed value is below the lower limit The relationship is linear, and the charging current curve is adjusted by a bidirectional converter.

[0093] The multi-source coordination control module also maintains a sliding time window to record the most recent A sequence of dynamic load assessment values ​​for each period. Distribution characteristics of evaluation values ​​within a statistical analysis window over a period of time: Calculate the mean. Standard deviation This will update the threshold range accordingly. The new upper limit value... Pick The new lower limit Pick The updated interval [ , Effective immediately. The threshold range update method enables the wind, solar, and energy storage system to automatically adjust its response sensitivity in response to changes in environmental conditions.

[0094] The pitch angle adjustment range of wind turbine generator sets is limited to [ , Within the specified range, if the calculated adjustment exceeds the mechanical limit, it will automatically be truncated to the nearest safe value. The tilt angle adjustment of the photovoltaic array is limited by structural strength; the angle of a single adjustment cannot exceed [a certain value]. The number of consecutive unidirectional adjustments does not exceed the set value. The rate of change of the energy storage unit's charging and discharging power is controlled within... Within the specified range, sudden power surges that could cause battery overheating are avoided. All constraint parameters are stored in the device characteristic database and are updated periodically with device maintenance records.

[0095] Wind turbine generator control commands are marked with the highest priority and transmitted via a dedicated real-time channel; photovoltaic array control commands are of medium priority and transmitted via a shared data bus; energy storage unit control commands are of normal priority and are allowed moderate delays. The transmission protocol includes a triple verification mechanism: the sending end attaches a CRC checksum when generating the command, and the receiving end performs arithmetic verification; key numerical fields use a one-inverse code dual-transmission mode; the command packet sequence number is strictly incremented, and a retransmission request is triggered if a packet is lost.

[0096] Wind turbine generator returns to actual pitch angle Rotation speed and output power Photovoltaic array feedback actual tilt angle Operating temperature and power generation current The energy storage unit reports the actual value of its charging and discharging power. State of charge and internal temperature The feedback data is compared with the expected adjustment target in a closed loop. If the deviation exceeds the allowable range, compensation adjustment is initiated. The compensation amount is calculated taking into account the equipment's response lag characteristics, and a recursive algorithm with a forgetting factor is used to gradually approximate the target value.

[0097] Each adjustment event records complete contextual information: environmental monitoring data at the trigger time, dynamic evaluation values, threshold range status, calculated theoretical adjustment amount, actual execution parameters, equipment feedback data, etc. The knowledge base supports multi-dimensional queries, allowing records to be filtered by time range, equipment type, adjustment direction, and other criteria. Regular data mining is performed on the knowledge base to identify high-frequency adjustment patterns and equipment performance degradation trends, providing a reference for preventative maintenance.

[0098] When the dynamic load assessment value continues When a cycle exceeds the pre-stored dynamic load assessment threshold range, the system upgrades to emergency response mode: automatic adjustment is suspended, and operations are only resumed after manual confirmation. In the event of device communication interruption, the latest valid instructions cached locally are executed cyclically. When sensor data is abnormal, alternative values ​​are estimated by referring to the operating status of similar devices. All abnormal events generate standardized reports, including exception codes, occurrence times, impact ranges, and handling measures, and alarms are pushed in real time through the operation and maintenance management platform.

[0099] By comparing real-time data streams from the environmental monitoring module with preset safety thresholds, a safety linkage module is established to trigger corresponding protection actions for three typical abnormal operating conditions: excessive wind speed, insufficient sunlight, and sudden load increases. A data sharing channel is established so that when an excessive wind speed alarm is received, the propeller pitch angle adjustment function is locked in advance. Load forecast information is exchanged with the external power grid dispatching system to optimize the charging and discharging strategy of the energy storage unit. The system also interfaces with a meteorological data platform to obtain future sunlight intensity forecasts to pre-adjust the photovoltaic array angle. The linkage interface adopts a publish-subscribe model, receiving only a subset of data relevant to itself, ensuring information isolation and security.

[0100] The wind turbine controller is deployed inside the turbine tower and connected to the central monitoring center via an industrial Ethernet network; the photovoltaic array controller is installed in the inverter cabinet and uses fiber optic communication; the energy storage unit control module is integrated into an external battery management system and exchanges data via a CAN bus. The main control computer is configured with dual-machine hot standby, so that the standby machine takes over control within a set time when the main unit fails. Critical adjustment parameters are dual-backed up in non-volatile memory, and automatically restore the most recent operating state after a restart.

[0101] At the end of each monitoring cycle, calculate the changes in key indicators before and after adjustment: changes in wind energy utilization rate. Changes in light energy conversion efficiency Changes in the charging and discharging efficiency of energy storage units These changes are correlated with dynamic load assessment values ​​to verify the effectiveness of coordinated control. Fluctuations in these changes within a continuous monitoring period, and their deviation from pre-stored dynamic load assessment threshold ranges, can trigger alarms from a safety linkage module that monitors the environmental monitoring module's data stream in real time, compares it with preset safety thresholds to address abnormal conditions such as excessive wind speed, insufficient sunlight, and sudden load increases. The assessment results are fed back to the logic unit responsible for assigning initial weights (set to 0.35, 0.35, and 0.30 respectively) to wind energy input balance, solar energy input balance, and load demand fluctuations; dynamically adjusting the weight ratios based on the coefficient of variation; and saving the weight configuration records. Furthermore, the logic unit can optimize weight allocation by combining the abnormal monitoring results from the safety linkage module, which monitors the environmental monitoring module's data stream in real time, compares it with preset safety thresholds to trigger protective actions such as locking the pitch angle when wind speed exceeds limits, switching to backup power when sunlight is insufficient, and activating peak compensation for energy storage when load increases. This influences the adjustment strategy for the next cycle. Long-term assessment data forms equipment regulation characteristic curves, providing a benchmark reference for parameter initialization at similar sites.

[0102] The central monitoring center synchronizes its clock with each sub-device via the PTP protocol, with time deviations controlled to the millisecond level. Adjustment commands are appended with precise timestamps, and the devices compensate for network transmission latency based on their local clocks. Hardware timers are used for critical control cycles to avoid timing jitter caused by software scheduling. Historical data records use a unified timestamp for easy analysis of causal relationships in later stages.

[0103] The control algorithm is encapsulated into independent functional blocks, and different versions of adjustment strategies can be loaded through configuration. The device interface layer is abstracted into a unified service bus; adding new device types only requires adapting the interface protocol. The policy engine supports dynamic loading of rule scripts, allowing the adjustment logic to be updated without restarting the system. Extended functions will be gradually migrated to the production environment after sandbox testing and verification.

[0104] Example 4: See Figure 4 The safety linkage module operates based on real-time comparison of preset safety thresholds with actual monitoring data. This module employs a three-tiered safety protection system, addressing three typical abnormal operating conditions: excessive wind speed, insufficient sunlight, and sudden load increases. The safety linkage module continuously monitors the real-time data stream from the environmental monitoring module. When any monitored parameter exceeds the safety threshold, the corresponding protection action is immediately triggered, simultaneously initiating a coordinated response from associated devices.

[0105] The implementation process for wind speed safety protection is as follows: preset two levels of wind speed thresholds, and a first-level warning value. and Level 2 hazard values When the environmental monitoring module detects real-time wind speed achieve When the wind speed increases, a warning signal is sent to the core control node, and a wind speed increase prompt is displayed on the human-machine interface. The core control node is responsible for centrally processing all safety event decision logic, receiving data streams from the real-time environmental monitoring module and comparing them with preset safety thresholds to trigger abnormal operating condition protection actions, and is equipped with a display prompt interface. At this time, the wind turbine enters a power-limited operation mode, and the blade pitch angle begins to gradually increase, but is not yet fully locked. When the wind speed continues to rise to... Upon activation, the safety linkage module immediately sends an emergency shutdown command, activating the wind turbine's hydraulic braking module, locking the blade pitch angle at its current position, and stopping the wind turbine's yaw module. Simultaneously, the energy storage unit receives a power compensation command, increasing its discharge power to compensate for the shortfall caused by the sudden reduction in wind power generation. The entire response process is completed in milliseconds, with the time from wind speed exceeding the limit to the completion of the actuator's action controlled within a set range. The actuator refers to the specific components used to execute the safety protection actions triggered by the safety linkage module, including: when the wind speed reaches a level two danger threshold, activating the wind turbine's hydraulic braking module, the device for locking the blade pitch angle, and the yaw module to stop wind operation; when there is insufficient continuous sunlight, executing a switching device to disconnect the photovoltaic array from the main grid and initiating a standby diesel generator set to start a preheating program; when the load suddenly increases, activating the energy storage unit's power regulation device with peak compensation function, and an adjustable load control device to temporarily reduce the power of air conditioning, lighting, etc.

[0106] Preset lower limit threshold of light intensity and duration threshold When the environmental monitoring module detects real-time light intensity Below When the light intensity is within a certain range, start the duration timer. If sunlight does not recover within the specified time, it is determined to be a continuous lack of sunlight, triggering a switch in the photovoltaic array's operating mode. The switching process involves three steps: first, disconnecting the photovoltaic array from the main grid and switching power supply from the energy storage unit through the inverter; second, initiating the preheating process of the backup diesel generator set; and finally, adjusting load priorities and temporarily shutting down non-essential electrical equipment. If the sunlight intensity is within... Rebound within a time period If the above is true, then the switch to the ready state will be cancelled.

[0107] Preset instantaneous load safety threshold and duration grading threshold , When the instantaneous value of the load's power demand exceeds Upon such an event, the peak compensation function of the energy storage unit is immediately activated. The energy storage unit automatically allocates compensation power according to the extent of the over-limit: if the over-limit is within a set ratio, compensation is provided solely by the energy storage unit; if the over-limit is significant, demand-side response measures are simultaneously initiated, temporarily reducing the power of adjustable loads such as air conditioning and lighting in a preset sequence. When the duration of the load over-limit reaches [a certain threshold], [the system will continue to provide compensation]. When this happens, a support request signal is sent to the power grid; upon reaching this point... When this occurs, the forced load shedding procedure will be initiated. See Table 1.

[0108] Table 1: The security event log table is used to store key information about all triggered events and includes the following fields.

[0109]

[0110] Before executing the lockout command, the following conditions must be verified for wind speed safety protection: whether the current output power of the wind turbine generator is lower than the set value; whether the wind speed sensor data is confirmed by at least two independent sensors; and whether the wind speed change trend conforms to physical laws. Before switching power supply modes, the following conditions must be verified for solar radiation safety protection: the energy storage unit's state of charge is sufficient to support the expected power supply duration; the backup diesel generator has sufficient fuel; and critical loads have been marked as protected objects. Before activating the compensation function, the following conditions must be verified for load surges: measurement errors are eliminated by comparing adjacent meter readings; the load type is checked to ensure it falls within the compensable range; and the current available power of the energy storage unit meets the requirements.

[0111] The security linkage module includes a master controller and slave controllers. The master controller centrally processes the decision-making logic for all security events, while slave controllers are distributed across various controlled devices. When the master controller detects a security event, it generates a control command package containing elements such as event level, response strategy, and execution time limit, and distributes it to the relevant slave controllers via a real-time network. Upon receiving the command, the slave controller immediately executes localized control and feeds back the execution result to the master controller. The master controller summarizes the response status of each device to form a complete event handling report. This architecture of collaborative security event processing by the master and slave controllers ensures both the uniformity of centralized decision-making and the timeliness of distributed execution.

[0112] After a wind speed safety incident is resolved, normal operation is not immediately restored. Instead, a three-stage transition is implemented: first, the mechanical lock on the pitch angle is released while maintaining power-limited operation; then, the power limitation is gradually reduced; and finally, maximum power point tracking control is fully restored. During the solar irradiance recovery process, a solar intensity stabilization monitoring period is implemented. Only when the real-time solar intensity consistently exceeds the recovery threshold for a set duration will the system gradually switch back to the main photovoltaic power supply mode. After a load surge event ends, the energy storage unit's compensation power decreases in a ramp-up manner to avoid backfeeding from the grid.

[0113] Weekly safety incident statistics reports are generated, calculating indicators such as the frequency of occurrence, average duration, and equipment response latency for various incidents. Monthly in-depth analysis is conducted to identify the spatiotemporal patterns of incident occurrences: statistically analyzing the distribution characteristics of incidents at different times; analyzing the relationship between wind speed exceeding limits and seasonal variations; and studying the correlation between sudden load increases and passenger station train timetables. The analysis results are used to optimize threshold settings and improve response strategies.

[0114] Wind speed monitoring is equipped with three independent sensors installed in different locations; sunlight monitoring uses a primary and backup sensor group, automatically switching working units periodically; load monitoring employs a dual-loop acquisition device, comparing the consistency of data from the two channels in real time. The control signal transmission channel is configured with a primary and backup dual network, automatically switching to the backup network when the primary network is interrupted. Critical control commands are stored in a "one primary, two backup" mode to ensure that any single point of failure does not affect the integrity of the commands.

[0115] When a security incident occurs, the human-machine interface of the central monitoring center, which is linked to security monitoring, automatically displays the corresponding handling flowchart, showing the current execution steps and the next contingency plan. Operators can manually intervene in the automatic handling process, but all intervention operations require double confirmation. During the incident handling process, the real-time status parameters of the affected equipment are continuously displayed, and key operation buttons are equipped with anti-accidental touch protection. Afterwards, the complete incident handling process can be replayed, supporting frame-by-frame analysis of the system response timing.

[0116] Regularly conduct wind speed exceedance simulation tests, generating virtual wind speed data using wind turbine test modes; conduct insufficient lighting drills, artificially creating low-light environments using shading devices; organize load surge stress tests, connecting a high-power load simulator to generate peak loads. Record various system response parameters during the testing process and compare them with design specifications. Test results are used to calibrate sensor accuracy and optimize control algorithms.

[0117] The security linkage module and all connecting modules use a dedicated communication protocol, and data frames include urgency level indicators. The highest priority command can interrupt regular data transmission and be directly inserted at the front of the message queue. Transmission latency is controlled to the millisecond level; critical commands are appended with timestamps and lifecycle markers, and commands not delivered within timeout periods automatically become invalid. The communication link maintains heartbeat monitoring; a connection failure is determined by the consecutive loss of a set number of heartbeat packets.

[0118] Newly integrated security monitoring devices only need to implement standard interface protocols to be integrated into the existing protection system. New security event types can be registered via configuration without modifying the core processing logic. Response policies support script-based customization, allowing adjustments to specific handling procedures based on different site characteristics. Extended functions will be gradually deployed through a canary release approach after simulation verification.

[0119] Example 5: See Figure 5 The workflow of the operation strategy optimization module starts after the preset time window ends. The time window length is set to run continuously for 720 hours (30 days), and the optimization program is triggered when the window ends. First, the complete historical operation dataset within the time window is extracted, including the raw wind speed sequence, light intensity sequence, and load power demand sequence collected by the environmental monitoring module, as well as the evaluation value sequence generated by the dynamic load assessment module. The data extraction scope covers the entire monitoring period, and each data point includes a millisecond-level timestamp and the device acquisition source identifier.

[0120] For wind speed data, calculate the hourly average wind speed and identify the intraday wind speed fluctuation cycle; analyze the maximum wind speed difference over three consecutive days and record the frequency of extreme values. For light intensity data, group and statistically analyze the light intensity distribution for each time period, calculate the correlation of light intensity on the same day within the same quarter, and mark the location and duration of periods of sustained low light intensity. For load electricity demand data, establish load curve templates for weekdays and rest days, identify the peak period length and peak interval patterns, and calculate the slope change characteristics of load abrupt events. All feature items are coded and stored by category, forming a feature vector matrix of time-series features.

[0121] The future operational trend prediction model is constructed based on the feature vector matrix of time-series characteristics. The model training process employs time series decomposition technology, breaking down historical data into trend components, periodic components, and random components. Trend component analysis uses the least squares method to fit the long-term change curve; periodic component analysis extracts the main fluctuation frequencies through Fourier transform; and random component analysis establishes an autoregressive moving average model. The parameters of the future operational trend prediction model are dynamically updated with new data, employing a rolling training mechanism. Each optimization retains the basic parameters from the previous training, only incrementally training on new time window data. The output of the future operational trend prediction model includes a wind speed change prediction curve for the next 168 hours (7 days), a solar radiation intensity distribution prediction map, and a load demand trend prediction table.

[0122] The reference value for the blade pitch angle of the wind turbine generator is modified based on the predicted wind speed distribution: when the prediction indicates sustained high wind speed, the reference value is increased; when the prediction indicates sustained low wind speed, the reference value is decreased; when the predicted wind speed is stable, the reference value remains unchanged. The reference value for the tilt angle of the photovoltaic array is adjusted according to the sunshine forecast: during the sunny weather forecast period, the reference value is tilted towards the optimal receiving angle; during the cloudy / rainy weather forecast period, the reference value tends towards a safe angle. The reference values ​​for the charging and discharging power of the energy storage unit are set according to the load forecast curve: during peak load forecast periods, the charging reference value is lowered and the discharging reference value is increased; during low load forecast periods, the opposite is true. The correction results form a new reference parameter table, replacing the original configuration. The reference value of the wind turbine blade pitch angle is one of the core operating parameters of the wind-solar-storage system. Together with the reference value of the photovoltaic array tilt angle, the reference value of the energy storage unit's charging and discharging power, and the pre-stored dynamic load assessment threshold, it constitutes the system's operating parameter system. After a preset time window, the operation strategy optimization module extracts the temporal characteristics of historical operating data such as wind speed, establishes a future operating trend prediction model based on these characteristics, and corrects the model accordingly. For example, it raises the reference value when predicting sustained high wind speeds, lowers the reference value when predicting sustained low wind speeds, and maintains the reference value when predicting stable wind speeds. It is also the basis for the multi-source coordinated control module to adjust the wind turbine blade pitch angle. The multi-source coordinated control module adjusts the pitch angle based on the comparison between the dynamic load assessment value and the pre-stored dynamic load assessment threshold range, and this adjustment is subject to the mechanical limit range of the blade pitch angle. , The constraint will automatically truncate the calculated adjustment amount to the nearest safe value if it exceeds the limit range.

[0123] The execution strategy optimization module's library is divided into three data areas: the basic record area stores the unprocessed raw execution data; the feature analysis area stores the extracted time-series feature vectors; and the adjustment record area stores the history of each parameter adjustment in detail. Each adjustment record includes elements such as the parameter value before adjustment, the parameter value after adjustment, the data markers used for adjustment, and the adjustment execution time. Each record is accompanied by a pointer to the feature extraction results, allowing users to trace back the data basis for adjustment decisions.

[0124] The operation strategy optimization module maintains equipment operation counters, recording the number of blade pitch angle adjustments, photovoltaic tilt angle adjustments, and charge / discharge power adjustments. When the cumulative value of any counter reaches a preset threshold (e.g., more than 50 adjustments in a single month), a parameter calibration command is immediately generated. Command priority is divided into three levels: a single device threshold triggers a level 3 command, a dual device threshold triggers a level 2 command, and a full device threshold triggers a level 1 command. The command includes a detailed list of overshooting devices and a statistical report on the overshoot magnitude.

[0125] Upon receiving a parameter calibration command, environmental data from the two most recent complete time windows are extracted from the historical operational feature database. The inter-window average wind speed difference rate, the overlap of light intensity distribution, and the correlation coefficient of the load curve are calculated separately. These three difference values ​​are weighted and fused into a comprehensive environmental change index. A seasonal adjustment factor is incorporated into the index calculation formula to eliminate the influence of natural climate change.

[0126] The adaptive correction method for the pre-stored dynamic load assessment threshold range is as follows: The comprehensive index of environmental changes is mapped to the threshold adjustment coefficient. For each unit increase in the index, the upper limit of the threshold range expands by a specified proportion, while the lower limit shrinks synchronously; if the index falls below the baseline value, the adjustment is reversed. The corrected threshold range must pass boundary checks: check that the new upper limit is not lower than the set safety limit; confirm that the new lower limit is not higher than the theoretical minimum value of dynamic load assessment; verify that the range width is within a preset reasonable range. The new threshold range that passes the check takes effect immediately, and the old range is transferred to the historical backup database.

[0127] During the feature extraction phase, if the data processing time exceeds the estimated time by 150%, a simplification mode is automatically activated: reducing the number of feature items and decreasing the analysis dimensions. During the model training phase, a resource usage warning line is set; the training process is terminated when memory usage exceeds the threshold. During the parameter correction phase, a mechanism to prevent misadjustment is implemented for the correction magnitude: a single correction exceeding 120% of the historical maximum correction magnitude requires secondary manual confirmation before it takes effect.

[0128] When the time window changes, an optimization start notification is pushed; the feature extraction progress is refreshed in real time in the form of a flowchart; the model prediction results are generated into a visual curve; the effect of parameter correction is marked with colors on the device control panel (red indicates significant changes, green indicates fine adjustments). Administrators can view historical optimization record comparison reports to analyze the long-term impact trend of each optimization on operating indicators.

[0129] Each optimization generates a configuration version number, which includes a timestamp and optimization type marker. The device controller stores the three most recent valid configuration versions, allowing rollback to the previous version in case of anomalies. Detailed version switching records include: operation time, rollback reason, and post-rollback monitoring data. The version comparison function supports displaying parameter differences between different versions side-by-side, using color blocks to indicate the magnitude of numerical changes.

[0130] After each optimization, continuous monitoring is conducted for 168 hours (7 days) to collect the deviation between actual operating data and predicted data. Verification indicators such as wind speed prediction error rate, light intensity matching degree, and load demand fit degree are calculated. Verification results are categorized into four levels: excellent, good, average, and poor, and stored in the optimization knowledge base as a reference for subsequent algorithm parameter adjustments. If three consecutive evaluations are rated as "poor," the model reconstruction process is automatically triggered.

[0131] The database is automatically backed up to an off-site storage node every morning at midnight. In the event of an optimization interruption, the most recent full backup point is checked, and the system is restored to its pre-interruption state. Critical compute nodes are configured with a checkpoint saving mechanism, saving intermediate results after each compute unit is completed. All data recovery operations require encrypted verification to prevent data tampering or loss.

[0132] An interface to a meteorological data platform is reserved, allowing the import of forecast data from professional meteorological agencies to enhance local models. An open power grid dispatch interface is provided to obtain regional load forecast data to supplement local analysis. A third-party algorithm import channel is set up, allowing the loading of optimization algorithm libraries to replace the core calculation module. External data input undergoes format conversion and reliability verification before being incorporated into the main optimization process.

[0133] The system learns from equipment maintenance history records and analyzes the distribution of routine maintenance periods. When a high-frequency maintenance period is detected, the end time of the time window is automatically delayed; compensatory optimization calculations are initiated immediately after maintenance activities. Equipment performance parameters collected during maintenance are directly integrated into the feature library to update the equipment state model, making the next round of optimization more consistent with the actual operating conditions of the equipment.

[0134] The logs record the following key information in detail: optimization start / end time, data extraction volume, feature analysis time, model training parameter version, baseline value correction list, threshold adjustment results, and anomaly handling events. The log files are stored using tamper-proof encoding and support quick location of key operation nodes by time indexing, meeting the requirements for traceability of operational quality.

[0135] It should be noted that, in this document, relational terms are used only to distinguish one operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0136] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A wind-solar-storage system applied to urban railway passenger stations, characterized in that, include: The environmental monitoring module is used to collect real-time data on wind speed, light intensity, and load power demand in the area where the city railway passenger station is located. The new energy input balance assessment module is used to receive the wind speed data, the light intensity data and the load power demand data, divide the monitoring period, and calculate the wind energy input balance, light energy input balance and load demand fluctuation within the preset standard range in each monitoring period. The dynamic load assessment module is used to generate dynamic load assessment values ​​for each monitoring period based on the wind energy input balance, solar energy input balance, and load demand fluctuation within each monitoring period. The dynamic load assessment module is equipped with a weight redistribution program, a configurable parameter table, and a historical weight record library. The wind energy input balance, solar energy input balance, and load demand fluctuation are used as three assessment indicators. Weight coefficients are assigned to the three assessment indicators, and the weight coefficient values ​​are stored in the configurable parameter table. When the data of a certain indicator deviates abnormally from the historical average within multiple consecutive monitoring periods, the weight allocation is automatically corrected according to a predetermined rule. The multi-source coordinated control module is used to adjust the blade pitch angle of the wind turbine generator, the tilt angle of the photovoltaic array, and the charging and discharging power of the energy storage unit based on the dynamic load assessment values ​​within each monitoring period. Specifically, it includes: The dynamic load assessment values ​​for each monitoring period are compared with the pre-stored dynamic load assessment threshold range; If the dynamic load assessment value is higher than the upper limit of the pre-stored dynamic load assessment threshold range during the current monitoring period, then the blade pitch angle of the wind turbine generator is increased, the tilt angle of the photovoltaic array is decreased, and the discharge power of the energy storage unit is increased. If the dynamic load assessment value is lower than the lower limit of the pre-stored dynamic load assessment threshold range during the current monitoring period, then the blade pitch angle of the wind turbine generator is reduced, the tilt angle of the photovoltaic array is increased, and the charging power of the energy storage unit is increased. The operation strategy optimization module is used to optimize the operation parameters based on historical operation data after the preset time window ends.

2. The wind-solar-storage system applied to urban railway passenger stations according to claim 1, characterized in that... The wind energy input balance is obtained through the following steps: Extract the maximum and minimum values ​​of the wind speed data within each monitoring period, and calculate the wind speed fluctuation difference; Obtain a preset standard wind speed fluctuation range, and combine the wind speed fluctuation difference with the standard wind speed fluctuation range to generate the wind energy input balance.

3. A wind-solar-storage system applied to a suburban railway passenger station according to claim 2, characterized in that... The light energy input equalization is obtained through the following steps: Extract the highest and lowest values ​​of the light intensity data within each monitoring period, and calculate the light intensity fluctuation difference; A preset standard light intensity fluctuation range is obtained, and the light energy input balance is generated by combining the light intensity fluctuation difference and the standard light intensity fluctuation range.

4. A wind-solar-storage system applied to a suburban railway passenger station according to claim 3, characterized in that... The load demand fluctuation is obtained through the following steps: Extract the peak and valley values ​​of the load power demand data within each monitoring period, and calculate the load demand fluctuation difference; Obtain a preset standard load fluctuation range, and combine the load demand fluctuation difference with the standard load fluctuation range to generate the load demand fluctuation degree.

5. A wind-solar-storage system applied to a suburban railway passenger station according to claim 4, characterized in that... The generation of dynamic load assessment values ​​for the wind-solar-storage system within each monitoring period specifically involves: The wind energy input balance, solar energy input balance, and load demand fluctuation within each monitoring period are weighted and calculated to output the dynamic load assessment value of the wind-solar-storage system.

6. A wind-solar-storage system applied to a suburban railway passenger station according to claim 5, characterized in that... It also includes a safety linkage module, used to perform the following operations: When the real-time wind speed collected by the environmental monitoring module exceeds the preset wind speed safety threshold, the blade pitch angle of the wind turbine generator set is locked. When the real-time light intensity collected by the environmental monitoring module is lower than the preset light intensity safety threshold, the photovoltaic array is switched to the backup power supply mode. When the instantaneous value of the load power demand collected by the environmental monitoring module exceeds the preset load safety threshold, the peak compensation function of the energy storage unit is activated.

7. A wind-solar-storage system applied to a suburban railway passenger station according to claim 6, characterized in that... The specific process of the optimization by the operation strategy optimization module includes: Extract the temporal characteristics of wind speed data, light intensity data, and load power demand data from the historical operating data; A future operational trend prediction model is established based on the aforementioned time-series characteristics; The reference values ​​for the blade pitch angle of the wind turbine generator, the tilt angle of the photovoltaic array, and the charging and discharging power of the energy storage unit are corrected based on the future operation trend prediction model.

8. A wind-solar-storage system applied to a suburban railway passenger station according to claim 7, characterized in that... The operation strategy optimization module also includes: A historical operation feature database is constructed to store the blade pitch angle adjustment records of the wind turbine generator, the tilt angle adjustment records of the photovoltaic array, and the charging and discharging power adjustment records of the energy storage unit; When the cumulative number of times the blade pitch angle adjustment record, tilt angle adjustment record, and charge / discharge power adjustment record reaches the preset adjustment count threshold, a parameter calibration command is triggered.

9. A wind-solar-storage system applied to a suburban railway passenger station according to claim 8, characterized in that... When the operation strategy optimization module responds to the parameter calibration command, it executes the following: Extract the environmental monitoring data difference values ​​of adjacent monitoring periods from the historical operation feature library; The pre-stored dynamic load assessment threshold range is adaptively corrected based on the differences in the environmental monitoring data.

Images