An energy coordination control system for hybrid energy storage system

Abstract

The present invention discloses an energy coordination control system for a hybrid energy storage system, comprising an energy storage device interface module, a data processing and analysis module, an energy coordination and distribution module, an energy management control module, a system operation and maintenance module, and a user interface module. The energy storage device interface module is used to connect and manage various energy storage devices, the data processing and analysis module is used to collect and process data, the energy coordination and distribution module is used to coordinate and distribute energy, the energy management control module is used to implement control strategies and monitor energy flow, the system operation and maintenance module is used to detect faults and perform maintenance, and the user interface module is used for user interaction with the system. The present invention discloses an energy coordination control system for a hybrid energy storage system, proposes an energy coordination algorithm based on model predictive control to predict and coordinate energy for a hybrid energy storage system, and proposes an intelligent control algorithm for a hybrid energy storage system based on a neural network to perform real-time detection and intelligent control of energy.

Description

Technical Field

[0001] The present invention relates to the fields of model predictive control, neural network and energy coordinated control, and in particular to an energy coordinated control system for a hybrid energy storage system. Background Art

[0002] Model predictive control technology is an advanced control strategy that aims to solve the energy prediction and coordination problems in hybrid energy storage systems. By establishing a system model and predicting future system behavior, model predictive control can optimize the distribution and use of energy, thereby improving the efficiency and stability of the system. Model predictive control technology is suitable for occasions that require precise control of energy flow and storage.

[0003] Neural network technology is an algorithm based on artificial intelligence that aims to solve the problems of real-time detection and intelligent control of energy. By simulating the neural network structure of the human brain, neural networks can handle complex nonlinear relationships and learn patterns and laws from large amounts of data. In hybrid energy storage systems, neural network technology can be used to monitor energy status in real time, predict system requirements, and respond quickly, thereby achieving smarter energy management.

[0004] However, the existing energy coordination control system for hybrid energy storage systems lacks the ability to accurately predict and coordinate energy flow, resulting in energy waste and insufficient supply. Secondly, the existing control strategy cannot effectively cope with complex system dynamics and changing external conditions, making the system slow to respond when faced with emergencies. In addition, the system lacks sufficient intelligence and flexibility and cannot be dynamically adjusted according to real-time data and prediction results, thus affecting the overall energy efficiency and stability. Summary of the invention

[0005] The purpose of the present invention is to provide an energy coordination control system for a hybrid energy storage system, so as to solve the problems that the existing energy coordination control system for a hybrid energy storage system proposed in the above background technology lacks the ability to accurately predict and coordinate energy flow, resulting in energy waste and insufficient supply, and the existing control strategy cannot effectively cope with complex system dynamics and changeable external conditions, making the system slow to respond when facing emergencies, and the system lacks sufficient intelligence and flexibility and cannot be dynamically adjusted according to real-time data and prediction results, thereby affecting the overall energy efficiency and stability.

[0006] To achieve the above-mentioned purpose, the present invention provides the following technical solutions: an energy coordination control system for a hybrid energy storage system, comprising an energy storage device interface module, a data processing and analysis module, an energy coordination and allocation module, an energy management control module, a system operation and maintenance module, and a user interface module, characterized in that: the energy storage device interface module is used to connect and manage the hybrid energy storage device to ensure that the device can effectively communicate with the system; the data processing and analysis module is used to collect, process and analyze data information from the hybrid energy storage device to support the decision-making and strategy implementation of subsequent modules; the energy coordination and allocation module includes an energy prediction coordination unit and an energy prediction coordination unit; the energy prediction coordination unit proposes an energy coordination algorithm based on model predictive control to predict future energy demand and the operating status of the energy storage device, so as to predict The energy optimization allocation unit is used to optimize the energy distribution among different energy storage devices according to real-time data and prediction results. The energy management control module includes an energy monitoring and detection unit and an energy intelligent control unit. The energy monitoring and detection unit is used to monitor and detect various parameters of the energy storage system in real time to ensure that the system operates in the best state. The energy intelligent control unit proposes a hybrid energy storage system intelligent control algorithm based on a neural network to intelligently control and adjust the energy flow and use strategy to ensure the efficient operation and stability of the system. The system operation and maintenance module is used to monitor the system status, detect faults and perform automatic maintenance to ensure the reliability and safety of the system. The user interface module is used to provide an interactive interface between the user and the system, display key parameters and support user configuration and management system.

[0007] Preferably, the energy storage device interface module connects to the hybrid energy storage device and realizes bidirectional transmission of data and control signals, thereby ensuring that the hybrid energy storage device can interact with the system in real time and respond to control instructions.

[0008] Preferably, the data processing and analysis module collects, processes and analyzes data from energy storage devices and sensors in real time, thereby ensuring that subsequent modules of the system can make accurate decisions and optimizations based on the latest information.

[0009] Preferably, the energy coordination and allocation module includes an energy prediction and coordination unit, which proposes an energy coordination algorithm based on model predictive control, and ensures that the system can adjust the energy coordination strategy in advance to cope with changes by predicting future energy demand and energy storage device status.

[0010] Specifically, the energy coordination algorithm based on model predictive control is as follows: First, an energy storage device model in the hybrid energy storage system is constructed. The energy storage device model takes into account the energy change of the energy storage device between consecutive time steps and the standby energy loss. The energy storage device model is used to accurately calculate the energy exchange amount at each time step, providing a basis for subsequent energy prediction and coordination. The specific formula of the constructed energy storage device model is expressed as follows:

[0011]

[0012] Among them, M α (k) represents the amount of power exchanged at time step k, α represents the energy storage device index, k represents the discrete time step, E α (k) is the amount of energy stored in the energy storage device α at discrete time step k, E α (k-1) represents the amount of energy stored in the energy storage device α at discrete time step k-1, Δt represents the time interval between time steps k and k-1, It is expressed as the amount of standby energy loss that the energy storage device α must cover at each time step k. Then, a prediction model is constructed by considering the system historical data, the current system status and future trends, including load demand, electricity price and renewable energy output. The specific formula is expressed as:

[0013]

[0014] Among them, P pred (k) is the predicted load, which represents the estimated value of future load at a specific time step k, P hist (k) represents the historical load, which is the load forecast calculated based on historical data and known patterns. ∈(k) represents the forecast error, which represents the uncertainty and forecast accuracy of the forecast model. By establishing an accurate forecast model, the system can make optimization decisions based on the forecast, thereby achieving more effective energy management and cost control. Secondly, by implementing the model predictive control strategy, the optimal coordination of energy flow in the hybrid energy storage system is achieved while considering the system dynamics and forecast uncertainty, optimizing the operation of the energy storage device and minimizing the operating cost, while ensuring the stable operation of the system under various uncertain conditions. The specific formula of the model predictive control strategy is expressed as:

[0015]

[0016] Among them, J MPC Expressed as a model predictive control function, c P (k+l) represents the electricity cost at time step k+l, p represents the electricity identifier, c E(k+l) represents the energy cost at time step k+l, E represents the energy identifier, P G (k+l) represents the power generation and load demand at time step k+l, G represents the load identifier, E stb (k+l) represents the energy level of the energy storage device at time step k+l, N represents the length of the prediction time horizon, l represents the index of the prediction step starting from the current time step k, i represents the index of the control variable, Ω represents the set of all control variables in the system, and c ∈ Denoted as the deviation penalty factor, c Sched Denoted as the scheduling penalty factor, ∈ i (k+l) 2 It is expressed as the prediction deviation of the control variable i at time step k+l, P Sched，i (k+l) represents the value of the control variable i calculated for pre-scheduling at time step k+l, P MPC，i (k+l)) 2 It is expressed as the value of the control variable i calculated by the model predictive control strategy at time step k+l. Then, online scheduling and real-time adjustment are used to deal with prediction errors and emergencies in actual operation to ensure that the system can be flexibly adjusted in real-time operation to adapt to the changing load and renewable energy input. The specific formula for online scheduling and real-time adjustment is expressed as:

[0017]

[0018] Among them, u MPC (k) represents the control input at time step k, is the variable value that makes the objective function take the minimum value, u(k) is the control input, and P G，Sched (k+l) represents the pre-dispatched generator setting, P G,MPC (k+l) represents the generator setting of the model predictive controller, which calculates the optimal control input u based on the current system state and prediction data. MPC (k), through the deviation from the pre-scheduled settings, and introduces a scheduling penalty factor c Sched To minimize the deviation, in this way, the system can be flexibly adjusted in real time to adapt to the changing load and renewable energy input, and finally,

[0019]

[0020] in,] SYS It is represented as the system implementation objective function, Φ is represented as the storage device set, j is represented as traversing each element in the set j, ∈ j (k) represents the storage device bias at time step k, It is represented as the energy value of the jth storage device in the model predictive control prediction, represents the energy value of the jth storage device in the actual operation of the system, Represents the value of the j-th control variable in the model predictive control prediction, It is represented by the value of the jth control variable in the actual operation of the system, C MPC Expressed as the control deviation penalty factor, the energy coordination algorithm based on model predictive control introduces a penalty factor in the system operation to minimize the deviation of the control variable from the pre-scheduling setting, ensuring the optimal control effect of the system in actual operation. In this way, the system can maintain an efficient and stable operating state under various complex operating conditions and achieve good energy coordination effect.

[0021] Preferably, the energy coordination and allocation module includes an energy optimization allocation unit, which ensures that the energy efficiency and performance of the system reach an optimal state by dynamically adjusting the distribution of energy among various energy storage devices.

[0022] Preferably, the energy management control module includes an energy monitoring and detection unit, which ensures that system anomalies and failures can be discovered and responded to in a timely manner by monitoring energy flow and equipment status in real time.

[0023] Preferably, the energy management control module includes an energy intelligent control unit, which proposes an intelligent control algorithm for the hybrid energy storage system based on a neural network, and intelligently controls and adjusts the energy flow and usage strategy through a neural network model to ensure the efficiency and stability of the system under various operating conditions.

[0024] Specifically, the intelligent control algorithm of the hybrid energy storage system based on neural network is as follows: First, a system model of the hybrid energy storage system is constructed to provide a basis for energy coordinated control. The model includes batteries, supercapacitors and bidirectional energy converters. The system model formula is specifically expressed as:

[0025] V j，oc =V j +I j R j

[0026] Among them, V j，oc It is represented as the open circuit voltage of the jth energy storage element, j is the energy storage element index, oc is the open circuit voltage identifier, V j It is expressed as the actual voltage of the jth energy storage element, which represents the voltage measured when current flows. j The current flowing through the jth energy storage element, R jThe internal resistance of the jth energy storage element is expressed as, and the voltage drop of the energy storage element under actual working conditions is calculated by constructing the system model of the hybrid energy storage system, taking into account the internal resistance R when the current flows through the element. j The voltage loss caused by the voltage loss helps to accurately predict and control the energy flow. Then, a model of a bidirectional energy converter is established. The converter works in Boost and Buck modes. The establishment of a bidirectional energy converter helps to understand the energy flow control in the hybrid energy storage system to avoid inefficient operation. The specific formulas of the model of the bidirectional energy converter in Boost and Buck modes are expressed as follows:

[0027]

[0028]

[0029] Among them, D boost It is expressed as the duty cycle in Boost mode, D buck Expressed as the duty cycle in Buck mode, V b Indicated as the voltage level of the battery in its current state, V uc It is represented by the voltage level of the supercapacitor in the current state, L represents the ability of the inductor to store magnetic energy, and R L Expressed as the resistance value of the inductor winding material, I L Expressed as the current flowing through the inductor, I dmd It represents the current required by the system at the current moment, b represents the battery identifier, and uc represents the supercapacitor identifier. Secondly, a multi-input and multi-output neural network architecture is constructed. The input includes the load demand current, the state of charge of the supercapacitor, the battery current at the previous moment, and the hybrid energy storage device operating efficiency parameters that affect the energy management decision. The output is the optimal control decision of the bidirectional energy converter. The optimal strategy is solved for the hybrid energy storage device drive cycle data set through the dynamic programming algorithm to generate the input and target data sets required for training the neural network. The specific formula is expressed as follows:

[0030] X(k)=[I dmd (k), SoC uc (k), I b，prev (k-1),δ(k)]

[0031] Among them, X(k) represents the neural network input vector at the kth time step, I dmd (k) represents the load demand current at the kth time step, SoC uc (k) represents the supercapacitor charge state at the kth time step, I b，prev(k) represents the battery current at the k-1th time step, δ(k) represents the operating efficiency parameter of the hybrid energy storage device at the kth time step. Finally, the neural network is designed and trained. The neural network training is performed by adjusting the weights and biases to minimize the error between the network output and the target data. The specific formula is expressed as follows:

[0032] w n+1 =w n -(J T J+λI) -1 J T e

[0033] e=ty

[0034] y=f(w T X+b)

[0035] Among them, w n+1 Represented as the n+1th layer network weight vector, w n is represented as the nth layer network weight vector, n is the neural network layer index, J is the Jacobian matrix, J T is the transposed matrix of the Jacobian matrix, T is the transposed operation, λ is the regularization parameter used to control the convergence speed of the algorithm, I is the identity matrix, e is the error vector, t is the target output vector, y is the actual output vector of the neural network, f(·) is the Sigmoid activation function, w T is represented as the weight vector transposed matrix, X is represented as the input data matrix, and b is represented as the bias vector. By training a neural network that can accurately simulate the optimal energy management strategy, intelligent control can be provided for the hybrid energy storage system. During the training process, the neural network continuously improves its prediction ability and decision-making accuracy by adjusting weights and biases. Finally, the trained neural network can provide precise energy management control and realize the intelligent control of the hybrid energy storage system by the energy coordination control system.

[0036] Preferably, the system operation and maintenance module ensures that the system can be quickly diagnosed and maintained when problems occur by continuously monitoring the system's operating status and automatically detecting faults, so as to maintain high reliability and safety of the system.

[0037] Preferably, the user interface module ensures that the user can conveniently view the system status, adjust settings, and effectively manage and operate the system by providing an intuitive user interface and interactive functions.

[0038] Compared with the prior art, the present invention has the following beneficial effects:

[0039] 1. The energy prediction and coordination unit proposes an energy coordination algorithm based on model predictive control. First, the algorithm accurately calculates the energy exchange amount at each time step by constructing an energy storage device model. The model takes into account the energy changes and standby energy losses between consecutive time steps, thus providing a solid foundation for subsequent energy prediction and coordination. This precise energy calculation ensures the efficiency and accuracy of the system in energy management. Secondly, by introducing a prediction model based on historical data and current system status, the model predictive control algorithm can effectively predict future load demand, energy, and renewable energy output. This prediction model not only improves the accuracy of load prediction, but also takes into account the impact of prediction errors, so that the system can make more optimized decisions under uncertain conditions. In this way, the system can better manage and allocate energy resources, reduce operating costs, and improve energy utilization efficiency. Secondly, the model predictive control strategy achieves the optimal coordinated operation of energy storage devices in the hybrid energy storage system by optimizing energy flow. This strategy takes into account system dynamics and prediction uncertainty. Under the condition of susceptibility, the objective function is used for optimization to ensure the stability and economy of the system under various operating conditions. This not only reduces the operating cost of the system, but also improves the overall efficiency and reliability of the system. In practical applications, the model predictive control algorithm can flexibly respond to prediction errors and emergencies in operation through online scheduling and real-time adjustment. This real-time adjustment capability enables the system to quickly respond to changes in load and renewable energy input, ensuring that the system still operates efficiently in a changing environment. Finally, the model predictive control algorithm introduces a penalty factor in the system operation to minimize the deviation between the control variables and the pre-scheduling settings, ensuring the optimal control effect of the system in actual operation. In this way, the system can maintain an efficient and stable operating state under various complex operating conditions and achieve good energy coordination effects. In summary, the energy coordination algorithm based on model predictive control has significant advantages in improving system energy management efficiency, reducing operating costs, enhancing system stability and responding to dynamic changes, and can provide strong support for the optimized operation of hybrid energy storage systems.

[0040] 2. Sequencing analysis and error correction unit A high-throughput sequencing error correction algorithm based on statistical analysis is proposed. First, a solid foundation for energy intelligent control is laid by constructing a system model of a hybrid energy storage system. The model includes batteries, supercapacitors and bidirectional energy converters. It can accurately calculate the voltage drop of energy storage components under actual working conditions and take into account the voltage loss caused by internal resistance when current flows through the components, thereby achieving accurate prediction and control of energy flow. In the model of the bidirectional energy converter, the system can work in two modes. Through detailed formula calculations, the system can accurately adjust the duty cycle to ensure that the energy conversion efficiency between batteries and supercapacitors is maximized and inefficient operation is avoided. This precise energy flow control improves the overall performance and stability of the hybrid energy storage system. In the design of the intelligent control algorithm, a multi-input and multi-output neural network architecture is used. The network uses load demand current, supercapacitor charge state, battery current at the previous moment, and energy management decision-making to influence parameters. The data is used as input, and the driving cycle data set of the hybrid energy storage device is optimized through a dynamic programming algorithm to generate the input and target data sets required for training the neural network. This method ensures the real-time and accuracy of control decisions. By designing and training the neural network, the system can minimize the error between the network output and the target data, thereby simulating the optimal energy management strategy. During the training process, the neural network continuously improves its prediction ability and decision-making accuracy by adjusting weights and biases. Finally, the trained neural network can provide precise energy management control and realize intelligent management for the hybrid energy storage system. In general, the intelligent control algorithm based on neural network has demonstrated excellent prediction and control capabilities in the hybrid energy storage system, significantly improving the energy utilization efficiency and operational stability of the system. This intelligent energy management method not only optimizes energy flow control, but also improves the system's response speed and decision-making accuracy, providing an efficient and reliable solution for modern hybrid energy storage systems. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] The invention is further illustrated by the accompanying drawings, but the embodiments in the accompanying drawings do not constitute any limitation to the invention. A person skilled in the art can obtain other drawings based on the following drawings without creative work.

[0042] Figure 1 It is a structural schematic diagram of the present invention. DETAILED DESCRIPTION

[0043] The following will be combined with the drawings in the embodiments 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.

[0044] See also Figure 1 The present invention provides an energy coordination control system for a hybrid energy storage system, including an energy storage device interface module, a data processing and analysis module, an energy coordination and allocation module, an energy management control module, a system operation and maintenance module, and a user interface module, characterized in that: the energy storage device interface module is used to connect and manage the hybrid energy storage device to ensure that the device can communicate effectively with the system; the data processing and analysis module is used to collect, process and analyze data information from the hybrid energy storage device to support the decision-making and strategy implementation of subsequent modules; the energy coordination and allocation module includes an energy prediction coordination unit and an energy prediction coordination unit; the energy prediction coordination unit proposes an energy coordination algorithm based on model predictive control to predict future energy demand and the operating status of the energy storage device, so as to conduct pre-capacity coordination. The energy optimization allocation unit is used to optimize the energy distribution among different energy storage devices according to the real-time data and prediction results. The energy management control module includes an energy monitoring and detection unit and an energy intelligent control unit. The energy monitoring and detection unit is used to monitor and detect various parameters of the energy storage system in real time to ensure that the system operates in the best state. The energy intelligent control unit proposes a hybrid energy storage system intelligent control algorithm based on a neural network to intelligently control and adjust the energy flow and usage strategy to ensure the efficient operation and stability of the system. The system operation and maintenance module is used to monitor the system status, detect faults and perform automatic maintenance to ensure the reliability and safety of the system. The user interface module is used to provide an interactive interface between the user and the system, display key parameters and support users to configure and manage the system.

[0045] See also Figure 1 Furthermore, the energy storage device interface module connects to the hybrid energy storage device and realizes bidirectional transmission of data and control signals, thereby ensuring that the hybrid energy storage device can interact with the system in real time and respond to control instructions.

[0046] See also Figure 1 ,Furthermore, the data processing and analysis module collects, processes and analyzes data from energy storage devices and sensors in real time, ,ensuring that subsequent modules of the system can make accurate decisions and optimizations based on the latest information.

[0047] See also Figure 1Furthermore, the energy coordination and allocation module includes an energy prediction and coordination unit, which proposes an energy coordination algorithm based on model predictive control, and ensures that the system can adjust the energy coordination strategy in advance to cope with changes by predicting future energy demand and energy storage device status.

[0048] See also Figure 1 Furthermore, the energy coordination algorithm based on model predictive control is as follows: First, an energy storage device model in a hybrid energy storage system is constructed. The energy storage device model takes into account the energy change of the energy storage device between consecutive time steps and the standby energy loss. The energy storage device model is used to accurately calculate the energy exchange amount at each time step, providing a basis for subsequent energy prediction and coordination. The specific formula of the constructed energy storage device model is expressed as follows:

[0049]

[0050] Among them, M α (k) represents the amount of power exchanged at time step k, α represents the energy storage device index, k represents the discrete time step, E α (k) is the amount of energy stored in the energy storage device α at discrete time step k, E α (k-1) represents the amount of energy stored in the energy storage device α at discrete time step k-1, Δt represents the time interval between time steps k and k-1, It is expressed as the amount of standby energy loss that the energy storage device α must cover at each time step k. Then, a prediction model is constructed by considering the system historical data, the current system status and future trends, including load demand, electricity price and renewable energy output. The specific formula is expressed as:

[0051]

[0052] Among them, P pred (k) is the predicted load, which represents the estimated value of future load at a specific time step k, P hist (k) represents the historical load, which is the load forecast calculated based on historical data and known patterns. ∈(k) represents the forecast error, which represents the uncertainty and forecast accuracy of the forecast model. By establishing an accurate forecast model, the system can make optimization decisions based on the forecast, thereby achieving more effective energy management and cost control. Secondly, by implementing the model predictive control strategy, the optimal coordination of energy flow in the hybrid energy storage system is achieved while considering the system dynamics and forecast uncertainty, optimizing the operation of the energy storage device and minimizing the operating cost, while ensuring the stable operation of the system under various uncertain conditions. The specific formula of the model predictive control strategy is expressed as:

[0053]

[0054] Among them, J MPC Expressed as a model predictive control function, c P (k+l) represents the electricity cost at time step k+l, p represents the electricity identifier, c E (k+l) represents the energy cost at time step k+l, E represents the energy identifier, P G (k+l) represents the power generation and load demand at time step k+l, G represents the load identifier, E stb (k+l) represents the energy level of the energy storage device at time step k+l, N represents the length of the prediction time horizon, l represents the index of the prediction step starting from the current time step k, i represents the index of the control variable, Ω represents the set of all control variables in the system, and c ∈ Denoted as the deviation penalty factor, c Sched Denoted as the scheduling penalty factor, ∈ i (k+l) 2 It is expressed as the prediction deviation of the control variable i at time step k+l, P Sched，i (k+l) represents the value of the control variable i calculated for pre-scheduling at time step k+l, P MPC，i (k+l)) 2 It is expressed as the value of the control variable i calculated by the model predictive control strategy at time step k+l. Then, online scheduling and real-time adjustment are used to deal with prediction errors and emergencies in actual operation to ensure that the system can be flexibly adjusted in real-time operation to adapt to the changing load and renewable energy input. The specific formula for online scheduling and real-time adjustment is expressed as:

[0055]

[0056] Among them, u MPC (k) represents the control input at time step k, is the variable value that makes the objective function take the minimum value, u(k) is the control input, and P G,Sched (k+l) represents the pre-dispatched generator setting, P G,MPC (k+l) represents the generator setting of the model predictive controller, which calculates the optimal control input u based on the current system state and prediction data. MPC (k), through the deviation from the pre-scheduled settings, and introduces a scheduling penalty factor C Sched To minimize the deviation, in this way, the system can be flexibly adjusted in real time to adapt to the changing load and renewable energy input, and finally,

[0057]

[0058] Among them, J SYS It is represented as the system implementation objective function, Φ is represented as the storage device set, j is represented as traversing each element in the set j, ∈ j (k) represents the storage device bias at time step k, It is represented as the energy value of the j-th storage device in the model predictive control prediction, represents the energy value of the jth storage device in the actual operation of the system, Represents the value of the j-th control variable in the model predictive control prediction, It is represented by the value of the jth control variable in the actual operation of the system, c MPC Expressed as the control deviation penalty factor, the energy coordination algorithm based on model predictive control introduces a penalty factor in the system operation to minimize the deviation of the control variable from the pre-scheduling setting, ensuring the optimal control effect of the system in actual operation. In this way, the system can maintain an efficient and stable operating state under various complex operating conditions and achieve good energy coordination effect.

[0059] See also Figure 1 Furthermore, the energy coordination and allocation module includes an energy optimization allocation unit, which dynamically adjusts the energy allocation among the energy storage devices to ensure that the energy efficiency and performance of the system are optimal.

[0060] See also Figure 1 Furthermore, the energy management control module includes an energy monitoring and detection unit, which ensures that system anomalies and failures can be discovered and responded to in a timely manner by monitoring energy flow and equipment status in real time.

[0061] Preferably, the energy management control module includes an energy intelligent control unit, which proposes an intelligent control algorithm for the hybrid energy storage system based on a neural network, and intelligently controls and adjusts the energy flow and usage strategy through a neural network model to ensure the efficiency and stability of the system under various operating conditions.

[0062] See also Figure 1 Furthermore, the intelligent control algorithm of the hybrid energy storage system based on neural network is as follows: First, a system model of the hybrid energy storage system is constructed to provide a basis for energy coordinated control. The model includes batteries, supercapacitors and bidirectional energy converters. The system model formula is specifically expressed as:

[0063] V j，oc =V j +I j R j

[0064] Among them, V j，oc It is represented as the open circuit voltage of the jth energy storage element, j is the energy storage element index, oc is the open circuit voltage identifier, V j It is expressed as the actual voltage of the jth energy storage element, which represents the voltage measured when current flows. j The current flowing through the jth energy storage element, R j The internal resistance of the jth energy storage element is expressed as, and the voltage drop of the energy storage element under actual working conditions is calculated by constructing the system model of the hybrid energy storage system, taking into account the internal resistance R when the current flows through the element. j The voltage loss caused by the voltage loss helps to accurately predict and control the energy flow. Then, a model of a bidirectional energy converter is established. The converter works in Boost and Buck modes. The establishment of a bidirectional energy converter helps to understand the energy flow control in the hybrid energy storage system to avoid inefficient operation. The specific formulas of the model of the bidirectional energy converter in Boost and Buck modes are expressed as follows:

[0065]

[0066]

[0067] Among them, D boost It is expressed as the duty cycle in Boost mode, D buck Expressed as the duty cycle in Buck mode, V b Indicated as the voltage level of the battery in its current state, V uc It is represented by the voltage level of the supercapacitor in the current state, L represents the ability of the inductor to store magnetic energy, and R L Expressed as the resistance value of the inductor winding material, I L Expressed as the current flowing through the inductor, I dmd It represents the current required by the system at the current moment, b represents the battery identifier, and uc represents the supercapacitor identifier. Secondly, a multi-input and multi-output neural network architecture is constructed. The input includes the load demand current, the state of charge of the supercapacitor, the battery current at the previous moment, and the hybrid energy storage device operating efficiency parameters that affect the energy management decision. The output is the optimal control decision of the bidirectional energy converter. The optimal strategy is solved for the hybrid energy storage device drive cycle data set through the dynamic programming algorithm to generate the input and target data sets required for training the neural network. The specific formula is expressed as follows:

[0068] X(k)=[I dmd (k), SoC uc (k), I b，prev (k-1),δ(k)]

[0069] Among them, X(k) represents the neural network input vector at the kth time step, I dmd (k) represents the load demand current at the kth time step, SoC uc (k) represents the supercapacitor charge state at the kth time step, I b，prev (k) represents the battery current at the k-1th time step, δ(k) represents the operating efficiency parameter of the hybrid energy storage device at the kth time step. Finally, the neural network is designed and trained. The neural network training is performed by adjusting the weights and biases to minimize the error between the network output and the target data. The specific formula is expressed as follows:

[0070] w n+1 =w n -(J T J+λI) -1 J T e

[0071] e=ty

[0072] y=f(w T X+b)

[0073] Among them, w n+1 Represented as the n+1th layer network weight vector, w n is represented as the nth layer network weight vector, n is the neural network layer index, J is the Jacobian matrix, J T is the transposed matrix of the Jacobian matrix, T is the transposed operation, λ is the regularization parameter used to control the convergence speed of the algorithm, I is the identity matrix, e is the error vector, t is the target output vector, y is the actual output vector of the neural network, f(·) is the Sigmoid activation function, w T is represented as the weight vector transposed matrix, X is represented as the input data matrix, and b is represented as the bias vector. By training a neural network that can accurately simulate the optimal energy management strategy, intelligent control can be provided for the hybrid energy storage system. During the training process, the neural network continuously improves its prediction ability and decision-making accuracy by adjusting weights and biases. Finally, the trained neural network can provide precise energy management control and realize the intelligent control of the hybrid energy storage system by the energy coordination control system.

[0074] See also Figure 1 ,Furthermore, the system operation and maintenance module ensures that the system can be quickly diagnosed and maintained when problems occur by ,continuously monitoring the system's operating status and automatically detecting ,faults, so as to maintain the high reliability and safety of the system.

[0075] See also Figure 1,Furthermore, the user interface module ensures that users can easily view the system status, adjust settings, and effectively manage and operate the system by providing an intuitive user interface and interactive functions.

[0076] When used specifically, first, the energy storage device interface module is used to connect and manage the hybrid energy storage device to ensure that the device can communicate effectively with the system. Then, the data processing and analysis module is used to collect, process and analyze data information from the hybrid energy storage device to support the decision-making and strategy implementation of subsequent modules. Secondly, the energy coordination and allocation module includes an energy prediction coordination unit and an energy prediction coordination unit. The energy prediction coordination unit proposes an energy coordination algorithm based on model predictive control to predict future energy demand and the operating status of energy storage equipment in order to perform pre-capacity coordination operations. The energy optimization allocation unit is used to optimize energy distribution among different energy storage devices based on real-time data and prediction results. Allocation, then, the energy management control module includes an energy monitoring and detection unit and an energy intelligent control unit. The energy monitoring and detection unit is used to monitor and detect various parameters of the energy storage system in real time to ensure that the system operates in the best state. The energy intelligent control unit proposes a hybrid energy storage system intelligent control algorithm based on a neural network to intelligently control and adjust the energy flow and usage strategy to ensure the efficient operation and stability of the system. Finally, the system operation and maintenance module is used to monitor the system status, detect faults and perform automatic maintenance to ensure the reliability and safety of the system. The user interface module is used to provide an interactive interface between the user and the system, display key parameters and support users to configure and manage the system.

[0077] Although the present invention has been described in detail with reference to the aforementioned embodiments, it is still possible for those skilled in the art to modify the technical solutions described in the aforementioned embodiments, or to make equivalent substitutions for some of the technical features therein. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. An energy coordination control system for a hybrid energy storage system, comprising an energy storage device interface module, a data processing and analysis module, an energy coordination and allocation module, an energy management control module, a system operation and maintenance module, and a user interface module, characterized in that: The energy storage device interface module is used to connect and manage the hybrid energy storage device to ensure that the device can communicate effectively with the system. The data processing and analysis module is used to collect, process and analyze data information from the hybrid energy storage device to support the decision-making and strategy implementation of subsequent modules. The energy coordination and allocation module includes an energy prediction and coordination unit and an energy optimization allocation unit. The energy prediction and coordination unit proposes an energy coordination algorithm based on model predictive control to predict future energy demand and the operating status of the energy storage device to perform pre-capacity coordination operations. The energy optimization allocation unit is used to optimize the distribution of energy between different energy storage devices based on real-time data and prediction results. The energy management control module includes an energy monitoring and detection unit and an energy intelligent control unit. The energy monitoring and detection unit is used to monitor and detect various parameters of the energy storage system in real time to ensure that the system operates in the best state. The energy intelligent control unit proposes a hybrid energy storage system intelligent control algorithm based on a neural network to intelligently control and adjust energy flow and use strategies to ensure efficient operation and stability of the system. The system operation and maintenance module is used to monitor the system status, detect faults and perform automatic maintenance to ensure the reliability and safety of the system. The user interface module is used to provide an interactive interface between the user and the system, display key parameters and support user configuration and management system; First, an energy storage device model in the hybrid energy storage system is constructed. The energy storage device model takes into account the energy changes of the energy storage device between consecutive time steps and the standby energy loss. The energy storage device model accurately calculates the energy exchange amount at each time step, providing a basis for subsequent energy prediction and coordination. The specific formula of the constructed energy storage device model is expressed as follows: Among them, M α (k) represents the amount of power exchanged at time step k, α represents the energy storage device index, k represents the discrete time step, E α (k) is the energy stored in the energy storage device α at discrete time step k, E α (k-1) represents the energy stored in the energy storage device α at discrete time step k-1, Δt represents the time interval between time steps k and k-1, It is expressed as the amount of standby energy loss that the energy storage device α must cover at each time step k. Then, a prediction model is constructed by considering the system historical data, the current system status and future trends, including load demand, electricity price and renewable energy output. The specific formula is expressed as: Among them, P pred (k) is the predicted load, which represents the estimated value of future load at a specific time step k, P hist (k) represents the historical load, which is the load forecast calculated based on historical data and known patterns. ∈(k) represents the forecast error, which represents the uncertainty and forecast accuracy of the forecast model. By establishing an accurate forecast model, the system can make optimization decisions based on the forecast, thereby achieving more effective energy management and cost control. Secondly, by implementing the model predictive control strategy, the optimal coordination of energy flow in the hybrid energy storage system is achieved while considering the system dynamics and forecast uncertainty, optimizing the operation of the energy storage device and minimizing the operating cost, while ensuring the stable operation of the system under various uncertain conditions. The specific formula of the model predictive control strategy is expressed as: Among them, J MPC Expressed as a model predictive control function, c P (k+1) represents the electricity cost at time step k+1, p represents the electricity identifier, c E (k+1) represents the energy cost at time step k+1, E represents the energy identifier, P G (k+1) represents the power generation and load demand at time step k+1, G represents the load identifier, E stb (k+1) represents the energy level of the energy storage device at time step k+1, N represents the length of the prediction time horizon, 1 represents the index of the prediction step starting from the current time step k, i represents the index of the control variable, Ω represents the set of all control variables in the system, and c ∈ Denoted as the deviation penalty factor, c Sched Denoted as the scheduling penalty factor, ∈ i (k+1) 2 It is expressed as the prediction deviation of the control variable i at time step k+1, P Sched,i (k+1) represents the value of the control variable i calculated for pre-scheduling at time step k+1, P MPC,i (k+1) represents the value of the control variable i calculated by the model predictive control strategy at time step k+1. Then, online scheduling and real-time adjustment are used to deal with prediction errors and emergencies in actual operation, ensuring that the system can be flexibly adjusted in real-time operation to adapt to the changing load and renewable energy input. The specific formula for online scheduling and real-time adjustment is expressed as: Among them, u MPC (k) represents the control input at time step k, is the variable value that makes the objective function take the minimum value, u(k) is the control input, and P G,Sched (k+1) represents the pre-dispatched generator setting, P G,MPC (k+1) represents the generator setting of the model predictive controller, which calculates the optimal control input u based on the current system state and prediction data. MPC (k), through the deviation from the pre-scheduled settings, and introduces a scheduling penalty factor c Sched To minimize the deviation, in this way, the system can be flexibly adjusted in real time to adapt to the changing load and renewable energy input, and finally, Among them, J SYS It is represented as the system implementation objective function, Φ is represented as the storage device set, j is represented as traversing each element in the set j, ∈ j (k) represents the storage device bias at time step k, It is represented as the energy value of the jth storage device in the model predictive control prediction, represents the energy value of the jth storage device in the actual operation of the system, Represents the value of the j-th control variable in the model predictive control prediction, It is represented by the value of the jth control variable in the actual operation of the system, c MPC Expressed as the control deviation penalty factor, the energy coordination algorithm based on model predictive control introduces a penalty factor in the system operation to minimize the deviation of the control variable from the pre-scheduling setting, ensuring the optimal control effect of the system in actual operation. In this way, the system can maintain an efficient and stable operating state under various complex operating conditions and achieve good energy coordination effect.

2. The energy coordination control system for a hybrid energy storage system according to claim 1, characterized in that: The energy storage device interface module connects to the hybrid energy storage device and realizes bidirectional transmission of data and control signals, thereby ensuring that the hybrid energy storage device can interact with the system in real time and respond to control instructions.

3. The energy coordination control system for a hybrid energy storage system according to claim 1, characterized in that: The data processing and analysis module collects, processes and analyzes data from energy storage devices and sensors in real time, ensuring that subsequent modules of the system can make accurate decisions and optimizations based on the latest information.

4. The energy coordination control system for a hybrid energy storage system according to claim 3, characterized in that: First, the system model of the hybrid energy storage system is constructed to provide a basis for energy coordinated control. The model includes batteries, supercapacitors and bidirectional energy converters. The system model formula is specifically expressed as: V j,oc =V j +I j R j Among them, V j,oc It is represented as the open circuit voltage of the jth energy storage element, j is the energy storage element index, oc is the open circuit voltage identifier, V j It is expressed as the actual voltage of the jth energy storage element, which represents the voltage measured when current flows. j The current flowing through the jth energy storage element, R j The internal resistance of the jth energy storage element is expressed as, and the voltage drop of the energy storage element under actual working conditions is calculated by constructing the system model of the hybrid energy storage system, taking into account the internal resistance R when the current flows through the element. j The voltage loss caused by the voltage loss helps to accurately predict and control the energy flow. Then, a model of a bidirectional energy converter is established. The converter works in Boost and Buck modes. The establishment of a bidirectional energy converter helps to understand the energy flow control in the hybrid energy storage system to avoid inefficient operation. The specific formulas of the model of the bidirectional energy converter in Boost and Buck modes are expressed as follows: Among them, D boost It is expressed as the duty cycle in Boost mode, D buck Expressed as the duty cycle in Buck mode, V b Indicated as the voltage level of the battery in its current state, V uc It is represented by the voltage level of the supercapacitor in the current state, L represents the ability of the inductor to store magnetic energy, and R L Expressed as the resistance value of the inductor winding material, I L Expressed as the current flowing through the inductor, I dmd It represents the current required by the system at the current moment, b represents the battery identifier, and uc represents the supercapacitor identifier. Secondly, a multi-input and multi-output neural network architecture is constructed. The input includes the load demand current, the state of charge of the supercapacitor, the battery current at the previous moment, and the hybrid energy storage device operating efficiency parameters that affect the energy management decision. The output is the optimal control decision of the bidirectional energy converter. The optimal strategy is solved for the hybrid energy storage device drive cycle data set through the dynamic programming algorithm to generate the input and target data sets required for training the neural network. The specific formula is expressed as follows: X(k)=[I dmd (k),SoC uc (k),I b,prev (k-1),δ(k)] Among them, X(k) represents the neural network input vector at the kth time step, I dmd (k) represents the load demand current at the kth time step, SoC uc (k) represents the supercapacitor charge state at the kth time step, I b,prev (k-1) represents the battery current at the k-1th time step, δ(k) represents the operating efficiency parameter of the hybrid energy storage device at the kth time step. Finally, the neural network is designed and trained. The neural network training is performed by adjusting the weights and biases to minimize the error between the network output and the target data. The specific formula is expressed as follows: w n+1 =w n -(J T J+λI) -1 J T e e=ty y=f(w T X+b) Among them, w n+1 Represented as the n+1th layer network weight vector, w n is represented as the nth layer network weight vector, n is the neural network layer index, J is the Jacobian matrix, J T is the transposed matrix of the Jacobian matrix, T is the transposed operation, λ is the regularization parameter used to control the convergence speed of the algorithm, I is the identity matrix, e is the error vector, t is the target output vector, y is the actual output vector of the neural network, f(·) is the Sigmoid activation function, w T is represented as the weight vector transposed matrix, X is represented as the input data matrix, and b is represented as the bias vector. By training a neural network that can accurately simulate the optimal energy management strategy, intelligent control can be provided for the hybrid energy storage system. During the training process, the neural network continuously improves its prediction ability and decision-making accuracy by adjusting weights and biases. Finally, the trained neural network can provide precise energy management control and realize the intelligent control of the hybrid energy storage system by the energy coordination control system.

5. The energy coordination control system for a hybrid energy storage system according to claim 1, characterized in that: The system operation and maintenance module ensures that the system can be quickly diagnosed and maintained when problems occur by continuously monitoring the system's operating status and automatically detecting faults, so as to maintain high reliability and safety of the system.

6. The energy coordination control system for a hybrid energy storage system according to claim 1, characterized in that: The user interface module ensures that users can easily view system status, adjust settings, and effectively manage and operate the system by providing an intuitive user interface and interactive functions.

Images