Energy-saving train auxiliary converter control method, device and storage medium

Abstract

The present application relates to the field of train auxiliary systems, and discloses an energy-saving train auxiliary converter control method, device and storage medium, including the following steps: S1, collecting the weight, vehicle type, ambient temperature, humidity, altitude and working time of the auxiliary converter of the train, and performing preprocessing; S2, calculating the environmental impact factor, vehicle load factor and historical energy consumption factor; S3, predicting the required power of the auxiliary converter through a machine learning model based on the calculated environmental impact factor, vehicle load factor and historical energy consumption factor; S4, determining the number of auxiliary converters that need to be put into operation based on the predicted required power and the basic principle of converter load rate; S5, combining the converter health prediction and working time factor, controlling the determined auxiliary converter to be put into operation. The energy-saving train auxiliary converter control method of the present invention realizes efficient use of energy through an intelligent control strategy, significantly improves energy efficiency and prolongs the life of the equipment.

Description

Technical Field

[0001] The present invention relates to the technical field of train auxiliary systems, and in particular to an energy-saving train auxiliary converter control method, device and storage medium. Background Art

[0002] In modern train transportation systems, the auxiliary converter module (ACM) is responsible for the auxiliary power supply of the entire train. The load includes air conditioning and ventilation systems, chargers and power consumption of various system controls, door control systems, braking systems, passenger information systems, etc. It is crucial to provide passenger comfort and ensure the effective operation of equipment. Traditional auxiliary converter control methods mainly rely on preset rules or simple feedback control for operation, usually running continuously in all weather conditions without considering the actual load conditions of the vehicle, environmental changes or energy consumption efficiency. This method leads to excessive energy consumption, especially under low load conditions, when all converters are put into operation and their transformers, filter circuits, etc. are basically running at no load, resulting in low energy efficiency.

[0003] In addition, fixed operation strategies cannot adapt to sudden environmental changes, such as sudden changes in temperature or a sharp change in the number of passengers, which require the auxiliary inverter to quickly adjust the output to maintain the comfort level of the cabin environment. Inflexible control not only increases energy consumption, but may also accelerate equipment wear and shorten its service life due to excessive equipment load. In addition, excessive use of inverters may also lead to unnecessary maintenance costs and downtime, affecting overall transportation efficiency and reliability.

[0004] Therefore, there is an urgent need to develop a more intelligent and efficient control strategy to respond to changes in the operating environment in real time, optimize energy consumption, extend equipment life, and improve passenger comfort. This strategy should be able to dynamically adjust the operation of the converter and optimize it according to actual needs, thereby overcoming the shortcomings of the existing technology. Summary of the invention

[0005] In view of the shortcomings of the prior art, the present invention provides an energy-saving train auxiliary inverter control method, device and storage medium to solve the problems of low energy efficiency, excessive equipment wear and inability to flexibly respond to load and environmental changes in traditional train auxiliary inverter control methods, optimize energy use through intelligent control strategies, and improve system responsiveness and equipment service life.

[0006] To achieve the above objectives, the present invention is implemented through the following technical solutions: A method for controlling an energy-saving train auxiliary converter comprises the following steps:

[0007] Collect the train weight, model, ambient temperature, humidity, altitude and auxiliary converter working hours, and perform pre-processing;

[0008] Calculate environmental impact factors, vehicle load factors, and historical energy consumption factors;

[0009] The required power of the auxiliary converter is predicted through a machine learning model based on the calculated environmental impact factors, vehicle load factors, and historical energy consumption factors;

[0010] Determine the auxiliary converters that need to be put into operation based on the predicted power demand and converter load factor;

[0011] The determined auxiliary converter is controlled to be put into operation by combining the converter health prediction and the working time factor.

[0012] Preferably, the environmental impact factor It is calculated by the following formula:

[0013]

[0014] in, , , is the pre-set weight coefficient, is the ambient temperature, For humidity, For altitude.

[0015] Preferably, the vehicle load factor It is calculated by the following formula:

[0016]

[0017] in, is the total weight of the current train, is the total weight of the train at maximum load.

[0018] Preferably, the historical energy consumption factor It is calculated by the following formula:

[0019]

[0020] in, and For the Environmental impact factors and load factors for each operation, For the Auxiliary converter power for each operation, is the number of runs.

[0021] Preferably, the method further comprises the step of determining the optimal number of auxiliary converters and operating time using a dynamic programming algorithm, and the specific steps include:

[0022] Defining states and decision variables , where the status Represents the system state at the current time step, including the current time, the number of auxiliary converters in operation, and the remaining required power; decision variables Represents the number of auxiliary converters that are put into operation or stopped in the current state;

[0023] Setting the initial state, including the initial time, the initial number of auxiliary converters and the initial required power;

[0024] Calculate the cost function for each state , considering the auxiliary converter operating cost, energy consumption and efficiency in meeting the required power;

[0025] Use recursive relations to calculate the optimal path from the initial state to the final state;

[0026] For each time step, the steps of calculating the optimal path from the initial state to the final state using the recursive relation are repeated until a preset running time is reached or all power requirements are met;

[0027] According to the calculated optimal path, the number of auxiliary converters that should be put into operation or stopped at each time step is output.

[0028] Preferably, the recursive relationship is defined by the following state transfer equation:

[0029]

[0030] in, For in time ,state The minimum total cost under To take action After that, from the state Transfer to state probability.

[0031] Preferably, the machine learning model is a random forest or gradient boosting tree algorithm.

[0032] Preferably, the method further comprises the step of using a reinforcement learning algorithm to adjust the environmental impact factors and control parameters in real time, and the specific steps include:

[0033] A reward function is defined, which calculates a reward value based on the energy efficiency of the auxiliary converter system, the operation stability of the equipment, and the degree to which a predetermined temperature control target is met;

[0034] Set the state space in the reinforcement learning algorithm, including the current environmental factors , vehicle load factor and historical energy consumption factor ;

[0035] Define the action space, including adjusting the number of auxiliary converters started and modifying the weight coefficients of environmental impact factors , and ;

[0036] Initialize the reinforcement learning model and set the learning rate and discount factor;

[0037] During the operation of the system, state data is collected in real time, and the optimal action is selected through the reinforcement learning model based on the current state and action space;

[0038] Execute the selected action, adjust the operating state of the auxiliary converter and the weight of the environmental influencing factor, and collect the new state and corresponding reward value after execution;

[0039] Use the collected new states and reward values ​​to update the reinforcement learning model and optimize the model to improve the efficiency and accuracy of future decisions.

[0040] The present invention also provides an energy-saving train auxiliary converter control device, comprising:

[0041] A data collection unit configured to collect vehicle weight, vehicle type, ambient temperature, humidity, altitude, and operating hours of the auxiliary converter;

[0042] A data processing unit configured to normalize and process outliers on the collected data;

[0043] a prediction unit configured to predict a required power of the auxiliary converter by using a machine learning model;

[0044] The control unit is configured to control the working state of the auxiliary converter according to the predicted required power.

[0045] The present invention also provides a computer device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, the above method is implemented.

[0046] The present invention also provides a storage medium on which a computer program is stored. When the computer program is executed by a processor, the method described above is implemented.

[0047] The present invention provides an energy-saving train auxiliary converter control method, device and storage medium. It has the following beneficial effects:

[0048] 1. The present invention can dynamically adjust the power demand according to the actual running state and environmental conditions of the train by intelligently controlling the start and stop of the auxiliary converter, thereby reducing unnecessary energy consumption. By optimizing the running time and number of converters, it ensures that the power consumption is minimized while meeting the comfort requirements, thus achieving efficient use of energy.

[0049] 2. The present invention optimizes the frequency of use and load distribution of the converter through an algorithm, avoiding frequent switching and overload operation of the equipment, thereby reducing the wear and failure rate of the equipment. In the long run, this control strategy can significantly extend the service life of the auxiliary converter and related equipment, and reduce maintenance and replacement costs.

[0050] 3. The present invention uses machine learning and reinforcement learning algorithms to analyze and predict the power requirements of the auxiliary converter in real time and make control decisions quickly and accurately. This highly automated control not only improves the operating efficiency of the system, but also enhances the ability to respond to sudden changes, such as sudden temperature changes or passenger flow fluctuations.

[0051] 4. The present invention accurately controls the power output of the inverter and adjusts the working state of the air conditioner and heating equipment according to the real-time temperature and humidity requirements in the car to ensure passenger comfort. This meticulous temperature control management not only improves passenger satisfaction, but also improves the overall quality of train services.

[0052] 5. The present invention helps to reduce carbon emissions during train operation by reducing energy consumption and optimizing equipment utilization efficiency.

[0053] 6. The energy-saving train auxiliary converter control method of the present invention realizes efficient use of energy through intelligent control strategies, significantly improves energy efficiency, prolongs equipment life, and optimizes passenger comfort. At the same time, the control parameters are adjusted in real time using machine learning and reinforcement learning algorithms, which enhances the response speed and accuracy of the system and reduces environmental impact. This comprehensive energy-saving solution not only improves train operation efficiency and passenger experience, but also meets the global needs of sustainable development and environmental protection, and has important practical and commercial value. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] Figure 1 A logical schematic diagram of the method of the present invention;

[0055] Figure 2 It is a schematic diagram of the train auxiliary power supply system of the present invention;

[0056] Figure 3 It is a schematic diagram of the method flow of the present invention;

[0057] Figure 4 It is a schematic diagram of the structure of the device of the present invention;

[0058] Figure 5 It is a schematic diagram of the computer device structure of the present invention.

[0059] Among them, 100, data acquisition unit; 200, data processing unit; 300, prediction unit; 400, control unit; 40, computer equipment; 41, processor; 42, memory; 43, storage medium. DETAILED DESCRIPTION

[0060] The following will be combined with the drawings in the specification of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0061] Please refer to the attached Figure 1 -Attached Figure 3 The embodiment of the present invention provides an energy-saving train auxiliary converter control method, which aims to optimize the operating efficiency of the train auxiliary converter through an intelligent control strategy, reduce energy consumption, and extend the service life of the equipment.

[0062] Specifically, the method comprises the following steps:

[0063] S1. Data collection and preprocessing

[0064] Step S1 mainly involves the collection and preprocessing of train operation related data, which is the basis for subsequent calculations and decisions. By accurately obtaining key parameters such as the train weight, vehicle type, ambient temperature, humidity, altitude, and auxiliary converter working hours, and effectively preprocessing these data, a reliable basis can be provided for the subsequent calculation of environmental impact factors, vehicle load factors, and historical energy consumption factors.

[0065] In this embodiment, step S1 is specifically as follows:

[0066] 1. Data collection:

[0067] Vehicle weight :The weight of the train is measured in real time by a weight sensor installed on the train chassis. The weight sensor continuously monitors the load condition of the train and transmits the data to the Train Control Management System (TCMS).

[0068] Model :The vehicle model information is usually preset when the train is manufactured and stored in the TCMS. This information can be obtained through the train configuration file or database query.

[0069] Ambient temperature :Install temperature sensors on the outside of the train to measure and record the ambient temperature in real time. The temperature sensor should have high sensitivity and accuracy to ensure the reliability of the data.

[0070] humidity :Humidity data is collected through external humidity sensors. Humidity sensors are usually used in conjunction with temperature sensors and are installed on the outside of the train to provide real-time environmental humidity information.

[0071] altitude : Altitude is measured by combining GPS system and air pressure sensor. GPS provides geographic location information, while air pressure sensor calculates altitude based on air pressure changes.

[0072] Auxiliary converter working time :Record the cumulative running time of each auxiliary converter. The timer inside the auxiliary converter records its working time in real time and feeds the data back to TCMS.

[0073] 2. Data preprocessing:

[0074] Normalization processing: In order to eliminate the impact of different data sources and dimensions, all collected data are normalized.

[0075] Outlier processing: Identify and exclude outliers through the triple standard deviation method to ensure the accuracy and reliability of the data.

[0076] Through data collection and preprocessing steps, we can ensure that high-quality basic data is obtained, providing a solid basis for subsequent calculations and decision-making.

[0077] S2. Calculate environmental impact factors, vehicle load factors and historical energy consumption factors

[0078] After completing the data collection and preprocessing of step S1, step S2 mainly calculates the environmental impact factor, vehicle load factor and historical energy consumption factor based on these preprocessed data. These factors will serve as important input parameters for the subsequent machine learning model to predict the required power. Through scientific and reasonable calculation methods, the energy consumption characteristics of the train under different operating conditions can be more accurately reflected, providing a basis for subsequent control decisions.

[0079] In this embodiment, step S2 is specifically as follows:

[0080] 1. Calculate environmental impact factors :

[0081] The environmental impact factor mainly reflects the impact of the external environment on the energy consumption of the train. It is calculated using the following formula:

[0082]

[0083] in, , , It is a pre-set weight coefficient determined based on historical data and expert experience, and is used to reflect the different degrees of impact of various environmental factors on energy consumption. is the ambient temperature, For humidity, For altitude.

[0084] 2. Calculate vehicle load factors :

[0085] The vehicle load factor reflects the ratio of the current train load to its maximum load and is calculated using the following formula:

[0086]

[0087] in, is the weight of the train, is the maximum load of the train. The higher the load factor, the heavier the train load, and the higher the demand for auxiliary converters may be.

[0088] 3. Calculate historical energy consumption factors :

[0089] The historical energy consumption factor combines past operating data to evaluate energy consumption patterns under different environmental and load conditions. The calculation formula is as follows:

[0090]

[0091] in, and For the Environmental impact factors and load factors for each operation, For the Auxiliary converter power for each operation, By calculating the historical energy consumption factor, we can capture the energy consumption characteristics in historical operation and provide a reference for predicting current demand.

[0092] Through the above steps, the environmental impact factor, vehicle load factor and historical energy consumption factor are accurately calculated, providing reliable input parameters for the demand power prediction in the subsequent step S3. The calculation method of these factors takes into account a variety of influencing factors and can more comprehensively reflect the impact of train operation status on energy consumption, thereby helping to improve the accuracy and effectiveness of the control strategy.

[0093] S3: Based on the calculated environmental impact factors, vehicle load factors, and historical energy consumption factors, the required power of the auxiliary converter is predicted through a machine learning model.

[0094] Step S3 mainly uses the environmental impact factor, vehicle load factor and historical energy consumption factor calculated in the above step S2 to predict the required power of the auxiliary converter through the machine learning model. This process requires combining a large amount of historical data to accurately predict future power demand through training models, thereby guiding the optimized operation of the auxiliary converter.

[0095] In this embodiment, step S3 is specifically as follows:

[0096] 1. Choose a machine learning model:

[0097] According to the characteristics of train operation data and the complexity of demand power prediction, select a suitable machine learning model for training. Commonly used models include ensemble learning algorithms such as Random Forest and Gradient Boosting Trees. These models can process high-dimensional data, capture complex nonlinear relationships, and have high prediction accuracy.

[0098] 2. Data preparation:

[0099] The environmental impact factor calculated in step S2 , vehicle load factor and historical energy consumption factor As input features, prepare the training dataset.

[0100] The training data set should include a large number of historical operation records, each of which contains the corresponding environmental impact factor, vehicle load factor, historical energy consumption factor and actual auxiliary converter required power. .

[0101] 3. Model training:

[0102] The selected machine learning model is trained using the prepared training data set. During the training process, the model learns the relationship between each input feature and the required power, and adjusts the internal parameters to minimize the prediction error.

[0103] In order to improve the generalization ability of the model, cross-validation technology is used to ensure that the model performs consistently on different data sets.

[0104] 4. Model verification and optimization:

[0105] The trained model is validated through the validation data set to evaluate its prediction performance. Common evaluation indicators include mean square error (MSE), mean absolute error (MAE), etc.

[0106] According to the verification results, the model parameters (such as the number and depth of trees) are adjusted to further optimize the model performance.

[0107] Real-time prediction:

[0108] In practical applications, the trained machine learning model is used to predict the impact of the current environmental factors. , vehicle load factor and historical energy consumption factor Perform real-time demand power forecasting.

[0109]

[0110] in, Represents a trained machine learning model function.

[0111] Through the above steps, the machine learning model is used to accurately predict the required power of the auxiliary converter, which provides a reliable basis for determining the number of converters put into operation in the subsequent step S4. This data-driven prediction method can fully consider multiple influencing factors, improve prediction accuracy, and thus optimize the operating efficiency and energy-saving effect of the auxiliary converter.

[0112] S4. Determine the auxiliary converters that need to be put into operation based on the predicted power demand and converter load rate.

[0113] Step S4 mainly determines the number of converters that need to be put into operation and selects specific converters for operation based on the auxiliary converter power demand predicted in step S3 and the basic principle of converter load rate. This step is the key to executing the control strategy. By reasonably scheduling the start and stop of the converter, energy consumption can be effectively reduced and the system operation efficiency can be improved.

[0114] In this embodiment, step S4 is specifically as follows:

[0115] 1. Define the load factor principle:

[0116] The load factor principle means that the auxiliary converter should operate within a certain percentage of its rated load to avoid underload or overload operation. The ideal load factor range is set at 50% to 75%. Within this range, the converter can operate with higher energy efficiency and reduce wear.

[0117] 2. The required power matches the number of converters:

[0118] According to the demand power forecast value obtained in step S3 , determine the number of auxiliary converters that need to be put into operation The determination of the number of converters requires consideration of the rated power of each converter. and current operating status.

[0119] The calculation formula is:

[0120]

[0121] in, It means rounding up to ensure that the required power is met, and Optimal Load Factor is the selected optimal load factor, such as 0.65 (or 65%).

[0122] 3.Converter status assessment:

[0123] Evaluate the current status of all auxiliary converters, including the accumulated operating hours , operating efficiency and health status. Through status assessment, the most suitable converter can be selected for operation to avoid overuse of some converters and their premature wear and tear.

[0124] The evaluation formula includes the following aspects:

[0125] Working hours weight:

[0126]

[0127] in, The maximum operating time allowed by the converter.

[0128] Health status weight:

[0129]

[0130] in, is the current detected converter loss value, is the preset loss threshold.

[0131] 4. Optimize the selection of converter:

[0132] According to the above weights and the required power, the optimal auxiliary converter combination is selected and put into operation. When selecting, converters with high operating efficiency and good health should be given priority.

[0133] The selection algorithm can use a heuristic algorithm or an optimization algorithm (such as dynamic programming, genetic algorithm, etc.) to solve the optimal converter combination solution.

[0134] 5. Control strategy output:

[0135] Output control strategy, including the specific converter number to be started, its running time and power allocation. The control strategy should be sent to each converter execution unit through the train control management system (TCMS) to ensure accurate execution.

[0136] Monitor the operating status of the converter in real time and make adjustments immediately if any abnormality is found (such as insufficient power or overload).

[0137] In this embodiment, through the above steps, the number of converters put into use can be dynamically adjusted according to actual needs and the status of the converters, and the optimal converter combination can be selected for operation, thereby maximizing the energy efficiency of the system and extending the service life of the equipment. This intelligent scheduling strategy effectively reduces energy waste and improves the overall performance of the train auxiliary converter system.

[0138] S5, combining the converter health prediction and the working time factor, controlling the determined auxiliary converter to be put into operation;

[0139] Step S5 mainly implements the previously determined auxiliary converter start-stop strategy to ensure that each converter works according to the predetermined operating time and power allocation. By combining the aforementioned converter health prediction and working time factor, the efficient operation of the auxiliary converter is ensured and its service life is extended. Through accurate control and real-time monitoring, effective management of the converter system can be achieved, further improving energy efficiency and extending equipment life.

[0140] In this embodiment, step S5 is specifically as follows:

[0141] 1. Control strategy execution:

[0142] According to the specific converter number, operation time and power distribution determined in step S4, a control output request instruction is sent to the train control management system (TCMS). The control output request instruction includes starting or stopping a specific converter and adjusting the output power of the converter.

[0143] Ensure that the control instructions are accurately transmitted to each converter execution unit and the converter operates according to the instructions.

[0144] 2. Real-time monitoring:

[0145] The TCMS monitors the operating status of each auxiliary converter in real time, including current output power, operating hours, temperature and other key parameters. The monitoring data should be fed back to the control center in a timely manner so that necessary adjustments can be made.

[0146] If an abnormality is detected during the monitoring process (such as insufficient power, overload or equipment failure), the system should immediately issue an alarm and take appropriate emergency measures (such as starting a backup converter or adjusting power distribution).

[0147] 3. Feedback adjustment:

[0148] Dynamically adjust the control strategy based on real-time monitoring data and operating results. Use feedback control algorithms (such as PID control) to fine-tune the operating parameters of the converter to ensure stable and efficient operation of the system.

[0149] Record operating data and update model parameters and control strategies to optimize future operating results.

[0150] 4. Regular maintenance and optimization:

[0151] Regularly inspect and maintain the converter equipment to ensure it is in good operating condition. Maintenance includes cleaning, lubrication, replacement of wearing parts and calibration of sensors.

[0152] Based on historical operation data and maintenance records, the control strategy and model parameters are optimized to improve the intelligence level and operation efficiency of the system.

[0153] In this embodiment, through the above steps, it is ensured that the auxiliary converter system can operate efficiently and stably according to the optimized control strategy. The real-time monitoring and feedback adjustment mechanism improves the response speed and adaptability of the system, ensuring that the best energy efficiency can be achieved under various operating conditions. This intelligent control method not only improves the overall performance of the system, but also extends the service life of the equipment, achieving the goals of energy saving and sustainable development.

[0154] The method implemented by the present invention can significantly improve the energy efficiency of the train auxiliary converter, reduce energy consumption, extend the life of the equipment, and automatically adjust the operating state through the intelligent control system to adapt to different environmental conditions and load requirements. This intelligent control strategy reduces human intervention and improves the convenience of operation and the response speed of the system.

[0155] In a preferred embodiment of the present invention, the method further comprises the step of determining the optimal number of auxiliary converters put into operation and the operating time using a dynamic programming algorithm to determine the optimal number of auxiliary converters put into operation and the operating time.

[0156] Through the dynamic programming algorithm, we can systematically consider the operating status of different time steps, find the overall optimal operating path, and ensure the maximum energy efficiency of the system. The specific steps are as follows:

[0157] 1. Define state and decision variables:

[0158] Status definition: Status Represents the system state at the current time step, including the current time 、Number of auxiliary converters put into operation and the remaining required power .

[0159] Decision variables: Decision variables Represents the number of auxiliary converters that are enabled or disabled in the current state. The value can be to start, stop or maintain the current number of converters.

[0160] 2. Set the initial state:

[0161] The initial state includes the initial time , Initial number of auxiliary converters and initial power requirement .

[0162] For example: , , Determined according to the predicted value in the previous step S3.

[0163] 3. Calculate the cost function for each state:

[0164] Cost Function Consider the operating cost, energy consumption and efficiency of the auxiliary converter in meeting the required power. The specific formula is:

[0165]

[0166] in, is the unit cost of converter operation, is the unit energy cost, is the penalty factor for unmet power demand.

[0167] 4. Use recursive relations to calculate the optimal path:

[0168] Use the recursive relation of dynamic programming to calculate the optimal path from the initial state to the final state. The recursive relation is as follows:

[0169]

[0170] in, For in time ,state The minimum total cost under To take action After that, from the state Transfer to state probability.

[0171] 5. Repeated calculation of the optimal path:

[0172] For each time step, the optimal path is calculated repeatedly using the recursive relation until the preset running time is reached. or meet all power requirements .

[0173] By updating the state and decision variables at each time step, the dynamic programming algorithm is able to continuously optimize the converter startup and shutdown decisions throughout the operation.

[0174] 6. Output the optimal path:

[0175] According to the optimal path calculated by dynamic programming, the number of auxiliary converters that should be put into operation or stopped at each time step is output. .

[0176] These outputs will be sent as control instructions to the Train Control Management System (TCMS) to ensure that the converter operates according to the optimal strategy.

[0177] This preferred embodiment further refines the auxiliary converter scheduling decision in step S4 through a dynamic programming algorithm. The demand power predicted in step S3 provides initial conditions and constraints for dynamic programming, and the dynamic programming algorithm optimizes the number of auxiliary converters and the operating time through the calculation of the recursive relationship and the cost function. This method can achieve real-time optimization in a complex operating environment, improve the energy efficiency and response speed of the system, and extend the service life of the equipment.

[0178] In a preferred embodiment of the present invention, the method further includes the step of using a reinforcement learning algorithm to adjust the environmental influencing factors and control parameters in real time. The aforementioned technical solution dynamically adjusts the operating strategy of the inverter through a reinforcement learning algorithm to achieve more efficient energy consumption management and equipment stability.

[0179] In this preferred embodiment, the specific steps of using the reinforcement learning algorithm to adjust the environmental impact factors and control parameters in real time are as follows:

[0180] 1. Define the reward function:

[0181] The reward function calculates the reward value based on the energy efficiency of the auxiliary converter system, the operating stability of the equipment, and the degree to which the predetermined temperature control target is met. The specific formula is as follows:

[0182]

[0183] in, , , are weight coefficients, which measure the importance of energy efficiency, operation stability and temperature control targets respectively.

[0184] 2. Set the state space:

[0185] The state space includes the current environmental factors , vehicle load factor and historical energy consumption factor . These state variables reflect the current operating status and historical data of the system.

[0186] 3. Define the action space:

[0187] The action space includes adjusting the number of auxiliary converters started and modifying the weight coefficients of environmental impact factors. , , . The specific actions can be expressed as:

[0188]

[0189] 4. Initialize the reinforcement learning model:

[0190] Choose a suitable reinforcement learning algorithm (such as Q-learning or Deep Q-Network, DQN), and set the learning rate and discount factor to balance short-term rewards and long-term benefits.

[0191] 5. Collect status data in real time:

[0192] During the operation of the system, real-time state data is collected, including ambient temperature, humidity, vehicle weight, and inverter operating status. These data are input into the reinforcement learning model as a representation of the current state.

[0193] 6. Choose the best action:

[0194] According to the current state and action space, the optimal action is selected through the reinforcement learning model. The ϵ-greedy strategy is used to balance exploration and exploitation to find the optimal solution.

[0195] 7. Perform the selected action:

[0196] Execute the actions selected by the model, adjust the operating status of the auxiliary converter and the weights of the environmental factors, monitor the system status after execution in real time, and collect the corresponding reward values.

[0197] 8. Update the reinforcement learning model:

[0198] Use the collected new states and reward values ​​to update the reinforcement learning model. Optimize model parameters through back-propagation or Q-value update algorithms to improve the efficiency and accuracy of future decisions.

[0199] This preferred embodiment further enhances the auxiliary converter scheduling decision in step S4 through the reinforcement learning algorithm. The demand power predicted in step S3 provides the initial conditions and constraints for the reinforcement learning, and the reinforcement learning algorithm optimizes the converter operation strategy in real time by dynamically adjusting the environmental influencing factors and control parameters. This method can achieve efficient energy consumption management and equipment stability in a complex and changeable operating environment.

[0200] The energy-saving train auxiliary converter control device described below and the energy-saving train auxiliary converter control method described above can refer to each other.

[0201] Please refer to the attached Figure 4The present invention also provides an energy-saving train auxiliary converter control device, comprising:

[0202] The data collection unit 100 is configured to collect vehicle weight, vehicle type, ambient temperature, humidity, altitude, and operating hours of the auxiliary converter;

[0203] The data processing unit 200 is configured to normalize and process outliers on the collected data;

[0204] A prediction unit 300 is configured to predict the required power of the auxiliary converter through a machine learning model;

[0205] The control unit 400 is configured to control the working state of the auxiliary converter according to the predicted required power.

[0206] The device of this embodiment can be used to execute the above method embodiment, and its principles and technical effects are similar, which will not be repeated here.

[0207] Please refer to the attached Figure 5 The present invention further provides a computer device 40, comprising: a processor 41 and a memory 42, wherein the memory 42 stores a computer program executable by the processor, and when the computer program is executed by the processor, the above method is executed.

[0208] The present invention further provides a storage medium 43 on which a computer program is stored. When the computer program is run by the processor 41, the above method is executed.

[0209] Among them, the storage medium 43 can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable red-only memory (PROM), read-only memory (ROM), magnetic storage, flash memory, disk or optical disk.

[0210] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for controlling an energy-saving train auxiliary converter, characterized in that: The following steps are involved: Collect the train weight, model, ambient temperature, humidity, altitude and auxiliary converter working hours, and perform pre-processing; Calculate environmental impact factors, vehicle load factors, and historical energy consumption factors; The required power of the auxiliary converter is predicted through a machine learning model based on the calculated environmental impact factors, vehicle load factors, and historical energy consumption factors; Determine the auxiliary converter that needs to be put into operation based on the predicted required power; Controlling the determined auxiliary converter to be put into operation; The environmental impact factors It is calculated by the following formula: ； in, is the pre-set weight coefficient, is the ambient temperature, For humidity, is the altitude; The vehicle load factor It is calculated by the following formula: ； in, is the total weight of the current train, is the total weight of the train at maximum load; The historical energy consumption factor It is calculated by the following formula: ； in, and For the Environmental impact factors and load factors for each operation, For the Auxiliary converter power for the second operation, is the number of runs.

2. The energy-saving train auxiliary converter control method according to claim 1, characterized in that: The method further includes the step of determining the optimal number of auxiliary converters and operating time using a dynamic programming algorithm, and the specific steps include: Defining states and decision variables , where the status Represents the system state at the current time step, including the current time, the number of auxiliary converters in operation, and the remaining required power; decision variables Represents the number of auxiliary converters that are put into operation or stopped in the current state; Setting the initial state, including the initial time, the initial number of auxiliary converters and the initial required power; Calculate the cost function for each state , considering the auxiliary converter operating cost, energy consumption and efficiency in meeting the required power; Use recursive relations to calculate the optimal path from the initial state to the final state; For each time step, the steps of calculating the optimal path from the initial state to the final state using the recursive relation are repeated until a preset running time is reached or all power requirements are met; According to the calculated optimal path, the number of auxiliary converters that should be put into operation or stopped at each time step is output.

3. The energy-saving train auxiliary converter control method according to claim 2, characterized in that: The recursive relationship is defined by the following state transition equation: ； in, For in time ,state The minimum total cost under To take action After that, from the state Transfer to state probability.

4. The energy-saving train auxiliary converter control method according to claim 1, characterized in that: The machine learning model is a random forest or gradient boosted tree algorithm.

5. The energy-saving train auxiliary converter control method according to claim 1, characterized in that: The method further includes the step of using a reinforcement learning algorithm to adjust the environmental impact factors and control parameters in real time, and the specific steps include: A reward function is defined, which calculates a reward value based on the energy efficiency of the auxiliary converter system, the operation stability of the equipment, and the degree to which a predetermined temperature control target is met; Set the state space in the reinforcement learning algorithm, including the current environmental factors , vehicle load factor and historical energy consumption factor ; Define the action space, including adjusting the number of auxiliary converters started and modifying the weight coefficients of environmental impact factors and ; Initialize the reinforcement learning model and set the learning rate and discount factor; During the operation of the system, state data is collected in real time, and the optimal action is selected through the reinforcement learning model based on the current state and action space; Execute the selected action, adjust the operating state of the auxiliary converter and the weight of the environmental influencing factor, and collect the new state and corresponding reward value after execution; Use the collected new states and reward values ​​to update the reinforcement learning model and optimize the model to improve the efficiency and accuracy of future decisions.

6. An energy-saving train auxiliary converter control device, used to execute the method according to any one of claims 1 to 5, characterized in that: include: A data collection unit configured to collect vehicle weight, vehicle type, ambient temperature, humidity, altitude, and operating hours of the auxiliary converter; A data processing unit configured to normalize and process outliers on the collected data; a prediction unit configured to predict a required power of the auxiliary converter by using a machine learning model; The control unit is configured to control the working state of the auxiliary converter according to the predicted required power.

7. A storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the method according to any one of claims 1 to 5 is implemented.

Images