Active power distribution network parallel control system with energy storage configuration

Abstract

The invention discloses an active power distribution network parallel control system containing energy storage configuration, and the parallel control system builds a virtual model of a power distribution network based on a multi-software joint real-time simulation platform, forms a virtual system capable of carrying out real-time operation, and carries out the real-time data interaction with an actual system. The parallel control system comprises a physical space module, a virtual system module, a communication system module and the like, so that the interaction control process of a virtual system and an actual system has real-time performance, real-time simulation can be carried out according to variable modification of the actual system in the system operation process, the real-time tracking capability is good, and the real-time performance of the system is improved. The real-time actual change of system parameters or the state of a control unit can be systematically carried out under the simulated working condition, and the control strategy is modified in combination with the actual working condition change. In addition, virtual system parameters can be modified according to actual changes of an actual system, control parameters or the state of a control unit are adjusted, and the real-time control effect is achieved.

Description

Technical Field

[0001] The present invention belongs to the technical field of distribution network energy storage for accessing distributed energy, and particularly relates to a parallel control system for an active distribution network with energy storage configuration built based on parallel control theory. Its construction and testing can both be based on the co-simulation platform of Simulink and Veristand. Background Art

[0002] Parallel control refers to the parallel interaction between the virtual and the real, that is, the parallel control strategy between the actual physical system and the artificial calculation. It takes into account the complexity of the actual project. By constructing a virtual system similar to the actual system, and through the information sharing and interaction between the virtual system and the actual system, a method for controlling the actual complex system is realized. Its core lies in the ACP theory, that is, artificial system, computational experiment, and parallel execution. Its specific implementation process is as follows: According to the specific system parameters and actual factors, construct an artificial system corresponding to the actual system, and by modifying the control unit or control parameters on the artificial system, after completing the computational experiment, through the interactive control between the artificial system and the actual system, the control optimization of the actual system is realized.

[0003] Based on the ACP theory, parallel control can be defined as a control method for optimizing tasks through the interaction between the virtual and the real. It combines the operation data of the virtual model and the actual system, and by comparing the states of the virtual model and the actual system in real time, the real-time monitoring and control of the system are realized. This technology uses technical means such as mathematical models, physical models, and data analysis to construct a virtual model corresponding to the actual system, and through learning and simulating the observed data of the actual system, the accurate prediction and simulation of the state, performance, and behavior of the actual system are realized.

[0004] Since the existing traditional simulation means generally adopt the way of data given, it is impossible to modify the unit parameters or the simulation model during the simulation operation, and it is impossible to show the real-time changes of the system. While the real-time simulator has the ability of fast calculation, and at the same time can interact with the actual signal through the I / O input and output board card, modify the model in real time according to the changes of the actual system or the actual control given by the user, show the dynamic changes of the system, and realize real-time simulation and real-time control.

[0005] In the field of electric power, as the part closest to the user load, the distribution network plays the role of distributing electric energy and is of extremely important significance to the social economic development and people's livelihood security. With the development of distributed generation technology, the access of various different distributed energy sources adds more randomness and uncertainty to the stable operation of the power grid. In addition, due to the different output characteristics of different distributed energy sources, the load of the distribution network is volatile, and the time-variability of the operating state of the active distribution network system increases significantly. With the rapid progress of material technology, the energy storage field has been further developed. Wide and flexible distributed energy storage is the main flexible controllable resource of the distribution network, which can effectively suppress the output power fluctuation of distributed energy sources and effectively improve the consumption capacity of distributed energy sources.

[0006] Therefore, how to reasonably and effectively configure the energy storage system in the distribution network and how to more optimally control the charging and discharging power of the energy storage system have become an urgent problem to be solved, with very important practical and theoretical significance. Therefore, a parallel control system for the distribution network can be established relying on a real-time simulation platform, and the control strategies under study can be verified and adjusted in real time in the parallel control system, and the actual power system can be guided and controlled. Summary of the Invention

[0007] The object of the present invention is a parallel control system for an active distribution network with energy storage configuration, which establishes an active distribution network model with energy storage configuration, and can issue control instructions according to the actual system changes or simulate specific working conditions for real-time simulation and real-time control.

[0008] Compared with the traditional off-line simulation method, the advantage of real-time simulation is that it can display the dynamic performance of the system in real time online. Thanks to the fast computing ability of the real-time simulator, and at the same time, using the input and output boards of the controller to establish a one-to-one mapping with the actual I / O signals, it can modify the parameters of the differential equations in the solution process in real time during the simulation, and reflect the external input or disturbance in the simulation of the model in real time. This process can be achieved by presetting signals or connecting to the actual I / O ports to realize the actual interaction process with external signals during the simulation. On this basis, real-time simulation can reduce the calculation step size to the microsecond level or even shorter on the basis of meeting the real-time requirements.

[0009] The parallel control system for an active distribution network with energy storage configuration includes:

[0010] A virtual system module, which establishes a virtual system of an active distribution network with energy storage configuration according to the operation state data of the actual system. The virtual system specifically includes an active distribution network simulation model with energy storage configuration and a real-time simulation platform;

[0011] The control optimization module designs an artificial interaction interface. Combining the algorithm optimization results, it issues control instructions to the virtual system through the interaction interface, and monitors the system status in real time to achieve dynamic modification and real-time feedback of the system.

[0012] In some embodiments, the construction of the active distribution network parallel control system with energy storage configuration is based on a multi-software simulation architecture suitable for the host computer management software of hardware-in-the-loop as the interaction basis. This multi-software simulation architecture serves as an intermediate bridge.

[0013] The active distribution network parallel control system with energy storage configuration includes a physical space module and an interactive communication system, where

[0014] The physical space module determines the system parameters and system structure according to the actual system, and issues control instructions to the virtual system according to different operating states or operating changes of the actual system.

[0015] The interactive communication system is used to connect the physical space module and the virtual system module, and perform data interaction to achieve real-time simulation and real-time control.

[0016] The virtual system module changes the control parameters or control units of the virtual system according to the system changes in the physical space, and conducts computational experiments in the virtual system.

[0017] In some embodiments,

[0018] The multi-software simulation architecture is a dual-computer architecture including a host computer and a slave computer. The host computer is connected to the slave computer, imports the required algorithms and models into the host computer, and then the host computer deploys the models to the slave computer to achieve data interaction between the host computer and the slave computer.

[0019] The host computer is a simulation software running on a PC, which tests and manages the system and provides experimental personnel with test control operations. Generally speaking, the host computer can be connected to the real-time simulation cabinet through connection methods such as a high-speed local area network, an Ethernet cable, and a bus for monitoring the operation of the slave computer.

[0020] The slave computer mainly provides the simulation of the controlled object. A real-time operating system is installed on its real-time hardware to ensure the real-time nature of the simulation. The behavioral models of the controlled object all run on the real-time operating system. Based on this platform, it is possible to quickly develop control algorithms and fault diagnosis-related functions for specific objects.

[0021] Specifically, the establishment of the real-time simulation architecture of the present invention realizes the co-simulation of Simulink and Veristand. Among them, Simulink serves as the host computer for designing the control model. After the processes such as debugging and compilation are error-free, the executable file can be deployed to the lower computer PXI controller through Veristand. Through the reserved input and output interfaces in the host computer model, that is, the I / O interfaces added in Simulink that can be recognized by Veristand, after establishing a one-to-one mapping data channel with the display interface, data interaction is carried out with the established simulation model to visually see the dynamic process of parameter changes in the interaction interface in real time.

[0022] Veristand includes one or more real-time execution targets, communicates with the host system through Ethernet, and provides an engine to communicate with subsequent processes. In the Veristand environment, the workspace window and the provided tools can be used to interact with the test system for data. The host computer system usually installs software tools such as MATLAB / Simulink for establishing the control algorithm model and compiling, as well as the Veristand real-time engine. Veristand provides real-time tracking tools to conveniently monitor the execution of the application, allows multiple data acquisition methods at the same time, and configures the I / O interface of the hardware through software.

[0023] The PXI controller, as the lower computer, is equipped with the real-time operating system NI Linux Real-Time, so it can quickly perform functions such as the execution of the algorithm model, the parameter assignment management of the I / O channel, and the real-time update of parameters. During the simulation process, through the transmission of data, the lower computer can receive the parameter values transmitted by the host computer in real time, and directly modify the differential equation in the solver according to the parameters in the next simulation step to achieve real-time changes in the simulation.

[0024] Based on the above-established real-time simulation architecture, a virtual system of the distribution network is further established. First, a distribution network model is established in Simulink. After configuring and compiling the model, interface settings are performed on the variables that need to interact with the outside world in the model. That is, after adding the corresponding interfaces in the Simulink library, compiler configuration and model compilation are carried out. Different from the offline simulation where different sub-models are used for different models, in this real-time simulation, all sub-modules adopt fixed-step simulation, and the target system file is set to the Linux compilation architecture, corresponding to the real-time operating system in the lower computer. After completion of the configuration, the executable file can be compiled and generated, and then the model is imported into Veristand. Through the preset interface parameters in the model, graphical UI processing is performed on it, and the model parameters can be modified in real time during the real-time simulation process, and the change of the active distribution network state with energy storage configuration caused by the modified parameters can be dynamically reflected.

[0025] Based on the deployment of the Veristand model, the present invention designs a human-machine interaction interface. This human-machine interaction interface can change the switch state during the simulation process, set the active and reactive output powers of the load and distributed energy sources. At the same time, relying on the high computing power of the real-time simulator, it can provide real-time feedback on the node voltage changes and the power of the distributed energy sources. During the actual operation process, according to the decision values given by the optimization algorithm, it can modify the active or reactive power output by the distributed generators (DGs) at different nodes of the distribution network in real time and monitor the changes in the state of the distribution network in real time, which is convenient for debugging control strategies and greatly improves the development efficiency. Through the verification of the voltage control strategy effect based on the virtual system, it can effectively guide the dispatching operation of the actual distribution network, thus realizing the core of the voltage parallel control of the active distribution network with energy storage configuration.

[0026] Corresponding to the parallel control theory, the present invention is equivalent to constructing a parallel control system for the actual distribution network. Among them, the artificial system is a virtual system built by using the real-time simulation platform constructed with Simulink and NI. The model built in this experiment is the IEEE 33-node system. As a commonly used example, this system is a classic distribution network model abstracted and equivalent from the actual system, and it is a typical radial network model. Based on the real-time simulation architecture, this model can be used to simulate and verify specific control units or proposed control strategies and guide the operation of the actual distribution network. Specifically, due to the real-time nature of the simulation architecture, during the operation of the virtual system, users can perform real-time control on the system according to the simulation requirements of specific working conditions of the actual distribution network or the control strategies proposed by themselves.

[0027] Since the capacity of the distributed energy sources connected to the distribution network is getting larger and larger, the distribution network has gradually developed towards high penetration. Because the distributed energy sources are easily affected by the outside world, it may cause voltage fluctuations at the nodes of the distribution network. In addition, due to the increase in penetration, the high-power output of DGs may generate reverse power flow, which may lead to serious voltage rise and threaten the complete and stable operation of the distribution network.

[0028] Therefore, in some embodiments, in the parallel control system of the active distribution network with energy storage configuration, the multi-objective optimal configuration of the distribution network energy storage system and the decision-making of the optimal charge and discharge power strategy based on the Archimedes optimization algorithm are applied. A multi-objective evaluation system with the minimum voltage fluctuation deviation of the distribution network, the minimum network loss consumption, and the optimal economy of energy storage planning is constructed. The established Archimedes optimization algorithm is used to solve the model, and finally, the optimal energy storage configuration scheme that can make the distribution network operate stably under the condition of frequent fluctuations in distributed power generation and the optimal charge and discharge power strategy during the operation cycle of the energy storage battery are obtained. Using the real-time simulation platform, according to the optimal control strategy output by the algorithm, the active or reactive output power of distributed energy is changed, and the change of the node voltage is monitored in real time. The implementation method of this strategy decision includes the following steps:

[0029] Step 1: Construct a distribution network and energy storage system model with access to distributed energy;

[0030] Step 2: Propose a specific multi-objective evaluation function;

[0031] Step 3: Apply the Archimedes optimization algorithm to optimize and solve the multi-objective model;

[0032] Step 4: Output the optimal control decision information, and the control decision includes the location and capacity determination decision information of the energy storage system and the optimal charge and discharge power strategy information of the energy storage system.

[0033] In some embodiments, the specific process of constructing the distribution network and energy storage system model with access to distributed energy in Step 1 includes the following steps:

[0034] Step (1-1): Construct a distribution network system operation model, mainly including the fluctuation deviation index of the distribution network node voltage and the line loss rate during the operation of the distribution network;

[0035] Step (1-2): Construct an energy storage system configuration model, mainly including: the cost required for the full-life operation cycle of the energy storage and the output model of the energy storage system accessing the high-penetration distribution network;

[0036] Step (1-3): Construct a high-penetration distribution network operation model, an energy storage system access distribution network configuration model, and constraint conditions. The constraint conditions specifically include: (1) the node voltage constraint and line current constraint of the high-penetration distribution network accessing distributed energy, (2) the power balance constraint, (3) the power flow constraint, (4) the energy storage battery capacity and energy multiple constraint, (5) the energy storage battery SOC state constraint, and (6) the distributed energy output constraint.

[0037] In some embodiments, the specific process of proposing a specific multi-objective evaluation function in Step 2 includes:

[0038] Step (2-1): Define the penalty objective function for the node voltage over-limit behavior for the system security of the distribution network nodes.

[0039] Step (2-2): Set the penalty objective function for the distribution network operation losses.

[0040] Step (2-3): Consider the location of the energy storage system in the distribution network and set the evaluation objective function.

[0041] In some embodiments, the specific process of applying the Archimedes optimization algorithm to optimize and solve the multi-objective model in Step 3 includes:

[0042] Step (3-1): Initialize the parameters of the Archimedes optimization algorithm and the optimization objective problem, and use the normalization method to process the objectives in the multi-objective comprehensive optimization process.

[0043] Step (3-2): Initialize the positions of all wooden block individuals, including all decision variables, and initialize and randomly assign the object volume, density, and acceleration during operation in water to each wooden block.

[0044] Step (3-3): Evaluate the initial population and select the individual with the best fitness value to assign the best position, best density, best volume, and best acceleration, and update and iterate the density and volume of each individual.

[0045] Step (3-4): Define the transfer operator and density factor, and gradually transform the algorithm from the search stage to the exploitation stage.

[0046] Step (3-5): In the exploration stage, update the acceleration of the wooden block individuals according to the mutual collision between the wooden block individuals; in the exploitation stage, there is no mutual collision between the wooden blocks, and use the "exploitation stage" formula to update the acceleration of the wooden block individuals.

[0047] Step (3-6): Normalize the acceleration and calculate its percentage change.

[0048] Step (3-7): Update the position and its affiliated attributes of each individual according to the transfer factor.

[0049] Step (3-8): Update the fitness value of each individual using the objective function, and record the individual with the best fitness value and its corresponding relevant attributes. Description of the Drawings

[0050] Figure 1 It is a schematic framework diagram of the active distribution network parallel control system with energy storage configuration in some embodiments of the present invention.

[0051] Figure 2 It is a flowchart for establishing and using the virtual system in some embodiments of the present invention.

[0052] Figure 3 The algorithm flowchart in some embodiments of the present invention.

[0053] Figures 4-7 The algorithm verification process in some embodiments of the present invention is specifically as follows:

[0054] Figure 4 They are the Whale Optimization Algorithm (WOA), Particle Swarm Optimization Algorithm (PSO), Genetic Algorithm (GA), and Archimedes Optimization Algorithm (AOA) respectively. Four optimization algorithms are used to solve the energy storage optimization configuration and energy storage power decision model. The current function curves with the number of iteration steps are as Figure 2 shown. The abscissa is the number of iteration steps, and the ordinate is the objective function value.

[0055] Figure 5 They are the full-time average amplitudes of the node voltages corresponding to the optimization configuration and control strategies of the four optimization algorithms. From top to bottom, they are the Archimedes algorithm (AOA), Particle Swarm Optimization Algorithm (PSO), Whale Optimization Algorithm (WOA), and Genetic Algorithm (GA).

[0056] Figure 6 and Figure 7 They are respectively the schematic diagrams of the decision output power and SOC state of the energy storage system under the Archimedes algorithm (AOA) within the full-time period cycle. Specific embodiments

[0057] The present invention will be further described below with reference to the accompanying drawings.

[0058] Combined with Figure 1 As shown in the figure, this embodiment proposes an active distribution network parallel control system with energy storage configuration. The basis for building the active distribution network parallel control system with energy storage configuration can be a joint simulation platform of Simulink and Veristand. It can be based on the NIPXI controller to establish an active distribution network model with energy storage configuration to form a virtual system, and can issue control commands according to the actual system changes or simulate specific working conditions for real-time simulation and real-time control. The specific working conditions can be the situation of simulating a specific model, including the structure and specific parameter quantities of the model, as well as the input and output situations. Because it is a simulation, it can be a simulation of an existing real system or a reproduction of a classic model.

[0059] Veristand can manage real-time tasks on the NIPXI controller, configure test sequences through a graphical interface, monitor input and output signals, record data, and perform stimulus generation and other test-related operations.

[0060] The active distribution network parallel control system with energy storage configuration includes:

[0061] The physical space module determines system parameters and system structure according to the actual system, and issues control instructions to the virtual system according to different operating states or operating changes of the actual system;

[0062] The virtual system module can establish an active distribution network virtual system with energy storage configuration according to the operating state data of the actual system. At the same time, it can also change the control parameters or control units of the virtual system according to the system changes in the physical space, and conduct computational experiments in the virtual system. The virtual system specifically includes an active distribution network simulation model with energy storage configuration and a real-time simulation platform;

[0063] The interactive communication system is used to connect the above two modules and conduct data interaction to achieve real-time simulation and real-time control;

[0064] The control optimization module designs an artificial interaction interface. Combining the algorithm optimization results, it issues control instructions to the virtual system through the interaction interface, and monitors the system state in real time to achieve dynamic modification and real-time feedback of the system.

[0065] In this embodiment, the virtual system module takes the establishment of an active distribution network virtual system with energy storage configuration according to the operating state data of the actual system as an example, such as: adopting the standard model of IEEE 33-node.

[0066] Combined with Figures 2 to 3 the content shown, the construction of the above virtual system module can specifically include the following steps:

[0067] Step A: Establish a distribution network digital model with high-penetration DG on the co-simulation platform, and proceed to Step B. In this embodiment, the IEEE 33-node distribution network model is selected as the standard model for studying the power system. The IEEE 33 distribution network model includes 33 nodes, 32 branches, and 5 tie-switch branches. Its power supply point rated voltage is 12.66 kV, the base power is 1 MW, the active load is 3751 kW, and the reactive load is 2300 kVar.

[0068] Step B: Use the power distribution network simulation model built in Simulink, compile it into a.dll file, import it into Veristand, and conduct real-time testing and monitoring on it. The specific implementation steps are as follows: After completing the establishment of the power distribution network model, configure the system environment, download the library NI Veristand Blocks related to Veristand in Simulink to provide an input-output interaction interface for both, and convert the model into a model supported by Veristand. Then, enter the compiler to compile the model. Depending on the system environment, two types of files, namely.so or.dll, will be generated at this time. Since the upper computer uses the Windows system, the compiled file generated here should be selected as the.dll file.

[0069] During the real-time simulation test of Veristand, the test objects are one or more. Each execution object is connected via an Ethernet cable to communicate with the upper computer system. The execution object in this case is a PXI controller, and the chassis controller model is PXIe-8880, with a built-in Linux real-time operating system. The execution object runs an engine that controls the system timing and is also used to control the communication between the system and the upper computer system. After its deployment, the upper and lower computers can perform real-time data interaction according to the instructions sent by the real-time working window.

[0070] Based on this communication foundation, after compiling and importing the established power model into Veristand, download it to the lower computer. With the real-time operating system carried by the lower computer PXI and its powerful computing ability, the upper computer can modify the instructions in real time and monitor its operating status simultaneously. By comparing the operating states of the power distribution network before and after parameter modification and the voltage control effects under different control strategies, the algorithm can be verified and optimized.

[0071] Step C: Verify the control unit or control algorithm based on the multi-software simulation architecture.

[0072] The present invention proposes a multi-objective optimization configuration and power decision method for a power distribution network energy storage system based on the Archimedes optimization algorithm to be applied to the above-mentioned active power distribution network parallel control system with energy storage configuration. Specifically, that is, using the Archimedes optimization method, a multi-objective evaluation system with the smallest voltage fluctuation deviation of the power distribution network, the smallest network loss consumption, and the best economy of energy storage planning is constructed. The Archimedes optimization algorithm is used to solve the model, and finally, the optimal energy storage configuration plan that can keep the power distribution network operating stably under the condition of frequent distributed power fluctuations and the best charge and discharge power strategy during the operation cycle of the energy storage battery are obtained. And through the real-time simulation platform, scheduling instructions are given according to the best optimization method given by the algorithm. The execution process of this algorithm includes the following steps:

[0073] Step 1: Build a distribution network and energy storage system model for accessing distributed energy, which specifically includes:

[0074] Step (1-1) Build a distribution network system operation model, which mainly includes the voltage fluctuation deviation index of the distribution network nodes and the line loss rate during the operation of the distribution network;

[0075] (1) Voltage fluctuation deviation index of distribution network nodes

[0076]

[0077] (2) Distribution network line loss rate

[0078]

[0079] In the formula, N is the number of nodes, M represents the number of branches; T is the statistical calculation period; U i,(t) , and U i,max are the voltage and reference voltage of the i-th node at time t respectively, and ΔU loss,ij(t) is the maximum allowable deviation; P

[0080] represents the network loss consumption of the ij branch at time t.

[0081] (1) Total life cycle cost of the energy storage system

[0082] f1 = C TCC + C OM - I sp - I sub

[0083]

[0084] In the formula: C TCC , C OM are the equivalent annual value investment cost and annual operation and maintenance cost of the energy storage system respectively; I sp , I sub are the annual low storage and high generation arbitrage and government subsidies for the sold electricity respectively; C inv is the fixed investment cost of a single energy storage system; N BESS is the installation quantity of the energy storage unit; P BESS,n , E BESS,n are the configured capacity and configured power of the n-th energy storage system respectively; a, b are the unit power cost and unit capacity cost of the energy storage system; μ CRF is the annual capital recovery rate; r is the discount rate; y is the service life; ρ on is the operation and maintenance coefficient.

[0085] (2) Energy storage system output model

[0086]

[0087] Where: P b,(t) is the output power of the energy storage unit at time t; λ d , λ c are the charge and discharge efficiencies of the energy storage battery respectively.

[0088] Step (1-3) constructs the operation model of the high-penetration distribution network, the configuration model of the energy storage system connected to the distribution network, and the constraint conditions. Specifically, it includes: (1) Node voltage constraints and line current constraints of the high-penetration distribution network with distributed energy access; (2) Power balance constraints; (3) Power flow constraints; (4) Energy storage battery capacity and energy multiple constraints; (5) Energy storage battery SOC state constraints; (6) Distributed energy output constraints.

[0089] Step 2: Propose a specific multi-objective evaluation function:

[0090] Step (2-1) defines the node voltage violation penalty objective function for the system security of the distribution network nodes;

[0091]

[0092] Step (2-2) formulates the penalty objective function for the distribution network operation loss;

[0093]

[0094] Step (2-3) defines the evaluation objective function of the capacity considering the position of the energy storage system in the distribution network:

[0095] F ALL = λ1f1 + λ2f2 + λ2f2

[0096]

[0097] Step 3: Apply the Archimedes optimization algorithm to optimize and solve the multi-objective model, specifically including:

[0098] Step (3-1) initializes the parameters of the Archimedes optimization algorithm and the optimization objective problem;

[0099] O i = lb i + rand × (ub i - lb i ); i = 1, 2,..., N

[0100] Where ub i , lb iThe upper and lower bounds of the search for each decision variable, respectively, O i represents the i-th population in the object.

[0101] Step (3-2) initializes the positions of all wooden block individuals, which includes all decision variables. And initialize and assign the object volume and density of each wooden block individual, and the acceleration during operation in water.

[0102]

[0103] Step (3-3) evaluates the initial population and selects the individual with the best fitness value and assigns it the best position x best , the best density den best , the best volume vol best , the best acceleration acc best . Update the density and volume of each individual iteratively.

[0104]

[0105] where den best , vol best are the best object density and volume found so far, respectively.

[0106] Step (3-4) defines the transfer operator and the density factor, and gradually transforms the algorithm from the search phase to the exploitation phase.

[0107]

[0108] where the transfer TF gradually increases with time until it reaches 1. Where t and t max are the current iteration number and the maximum iteration number, respectively. Similarly, the density decreasing factor d also helps the Archimedes algorithm to perform global to local search. When using the density factor, it decreases over time.

[0109] Step (3-5) utilizes the mutual collision between wooden block individuals in the exploration phase to update the acceleration of the wooden block individuals; in the exploitation phase, there is no mutual collision between the wooden blocks, and the "exploitation phase" formula is used to update the acceleration of the wooden block individuals.

[0110] (1) Exploration phase: If TF ≤ 0.5, collisions occur between objects, and randomly select a material to update the acceleration of the object at iteration T+1:

[0111]

[0112] (2) Exploitation phase: If TF > 0.5, there is no collision between objects, and the following formula is used to update the acceleration of the object at iteration t+1:

[0113]

[0114] Among them, acc mr , den mr , vol mr are the acceleration, density, and volume corresponding to a random material.

[0115] Step (3 - 6) normalizes the acceleration and calculates its percentage change.

[0116]

[0117] Among them, u and l are the normalization ranges and are set to 0.9 and 0.1 respectively. The norm determines the percentage of the step size by which each block will change. If block i is far from the global optimum, the acceleration value will be high - which means the object will be in the exploration phase; otherwise, in the exploitation phase. This shows how the search transitions from the exploration phase to the exploitation phase. Normally, the acceleration factor starts from a large value and decreases over time. This helps the search to move towards the global best solution while staying away from local solutions. However, it is worth mentioning that there may still be some search agents that need to stay in the exploration phase for more time than normal. Therefore, the Archimedes algorithm achieves a balance between exploitation and exploration.

[0118] Step (3 - 7) updates the position of each individual and its affiliated attributes according to the transfer factor.

[0119]

[0120] Among them, C1 and C2 are constant values 2 and 6 respectively.

[0121] Step (3 - 8) updates the fitness value of each individual using the objective function, records the individual with the best fitness value and its corresponding related attributes, and repeats the iterative loop until the convergence criterion is reached.

[0122] In the Archimedes optimization algorithm, an individual represents a solution in the problem space, and the related attributes describe the specific characteristics of the solution, such as: position, density, acceleration.

[0123] Position: Represents the coordinates of an individual in the search space. For the energy storage system configuration, it is the energy storage capacity and power level of each node.

[0124] Density: Simulates the density of an object in water and is used to affect the behavior of an individual (exploration or exploitation phase). In practical applications, it can be mapped to the cost - benefit ratio of the energy storage system or other similar concepts.

[0125] Volume: The volume of the simulated object, which can be associated with cost or performance metrics during the optimization process, such as the total capacity of the energy storage system.

[0126] Acceleration: It reflects the movement trend of an individual during the iteration process, similar to the learning rate or step size adjustment, and is used to guide how an individual updates its position.

[0127] In each iteration, the algorithm evaluates the fitness of all individuals according to the set objective function and selects the individual with the highest fitness (the lowest fitness for minimization problems) as the current best individual. Then, the algorithm records the position and other relevant attributes of this best individual and uses this information to guide the subsequent search process. As the number of iterations increases, the individuals in the population gradually converge to the global optimal solution until the preset stopping conditions are met, such as the maximum number of iterations or the change in the fitness value being less than a certain threshold.

[0128] In this way, AOA can effectively find the optimal energy storage configuration plan and charge-discharge strategy for stable operation of the distribution network in a complex multi-dimensional search space.

[0129] Step (4): Output the optimal control decision information.

[0130] Step (4-1): Output the decision information on the location and capacity of the energy storage system. The capacity decision information includes the location node, that is, at which node to introduce the energy storage unit.

[0131] Step (4-2): Output the optimal charge-discharge power strategy information of the energy storage system.

[0132] Figures 4 to 7 They are the Whale Optimization Algorithm (WOA), Particle Swarm Optimization Algorithm (PSO), Genetic Algorithm (GA), and Archimedes Optimization Algorithm (AOA). Four optimization algorithms are used to solve the energy storage optimization configuration and energy storage power decision-making model. The current function curves with the number of iteration steps are as Figure 4 shown. It can be seen from it that compared with other algorithms, the Archimedes Optimization Algorithm can quickly and effectively converge to a better objective value.

[0133] Figures 4 to 7 They are the full-time average amplitudes of the node voltages corresponding to the optimization configuration and power control strategies of the four optimization algorithms. From top to bottom, they are the Archimedes Algorithm (AOA), Particle Swarm Optimization Algorithm (PSO), Whale Optimization Algorithm (WOA), and Genetic Algorithm (GA). As shown in the figure, the voltage distribution curve determined by the Archimedes Optimization Algorithm is smoother and closer to the standard value compared with other algorithms. The configuration plans of the energy storage systems in the active distribution network obtained by solving with the four algorithms are shown in Table 1 below.

[0134] Table 1 Optimal Configuration Plan of Energy Storage System

[0135]

[0136] Figure 5 and Figure 6 are respectively schematic diagrams of the decision output power and SOC state of the energy storage system under the Archimedes algorithm (AOA) within the full-time period cycle.

[0137] In fact, as a parallel control system, users can adopt different control models and different control algorithms according to actual needs, modify and adjust them, and test and monitor both. Its advantage lies in that with the high computing power of the real-time simulator, it can provide real-time feedback on the node voltage changes and the power of distributed energy, and can also modify control parameters online. For example, it can control the access of specific energy storage units or the magnitude of their active and reactive power outputs, improving the test efficiency and achieving good test and monitoring effects. The focus of this invention is to utilize the superiority of the parallel control system, that is, to use the virtual system to modify and monitor relevant control algorithms and control units in real time, and after verifying and optimizing the algorithms based on the virtual system, it can guide the dispatching operation of the actual distribution network. In this invention, only case analysis verification is carried out based on the virtual system, and the verified control strategy is not used for actual system control.

[0138] The above undisclosed matters can all be achieved by existing technologies, so they will not be elaborated here.

[0139] The above example analysis is a case analysis of the content protected by the patent, and does not impose any formal restrictions on the present invention. Any modifications, equivalent replacements, improvements, etc. made to the above embodiments based on the technical essence of the present invention without departing from the technical solution of the present invention still fall within the scope of the technical solution of the present invention.

Claims

1. An active distribution network parallel control system with an energy storage configuration, characterized in that, It includes: A virtual system module that establishes an active distribution network virtual system with energy storage configuration based on the operation status data of the actual system. The virtual system specifically includes an active distribution network simulation model with energy storage configuration and a real-time simulation platform; A control optimization module that designs a human-computer interaction interface, combines the algorithm optimization results, issues control instructions to the virtual system through the interaction interface, and monitors the system status in real time to achieve dynamic modification and real-time feedback of the system.

2. The active distribution network parallel control system with an energy storage configuration according to claim 1, characterized in that The construction of the active distribution network parallel control system with energy storage configuration is based on a multi-software simulation architecture suitable for the host computer management software of hardware-in-the-loop as the interaction foundation, and this multi-software simulation architecture serves as an intermediate bridge; The active distribution network parallel control system with energy storage configuration includes a physical space module and an interactive communication system, where The physical space module determines the system parameters and system structure according to the actual system, and issues control instructions to the virtual system according to different operation states or operation change situations of the actual system; The interactive communication system is used to connect the physical space module and the virtual system module, and perform data interaction to achieve real-time simulation and real-time control; The virtual system module changes the control parameters or control units of the virtual system according to the system changes in the physical space, and conducts computational experiments in the virtual system.

3. The active distribution network parallel control system with energy storage configuration according to claim 2, wherein The multi-software simulation architecture is a dual-machine architecture including a host computer and a slave computer. The host computer is connected to the slave computer, imports the required algorithms and models into the host computer, and then the host computer deploys the models to the slave computer to achieve data interaction between the host computer and the slave computer; The host computer is a simulation software running on a PC, tests and manages the system, and provides test control operations for experimental personnel. The host computer is connected to the real-time simulation cabinet and is used to monitor the operation of the slave computer; The slave computer is used to provide the simulation of the controlled object. A real-time operating system is installed on its real-time hardware, and the behavior models of the controlled object all run on the real-time operating system.

4. The active distribution network parallel control system with an energy storage configuration according to claim 2, characterized in that, Based on the multi-software simulation architecture, a distribution network virtual system is built according to the actual power system for verifying and optimizing the control unit or control strategy. The specific building process includes the following steps: Step 1: Build a distribution network simulation model, configure and compile it, and then manually add an interface library adapted to the compilation framework and add an input / output interaction module to complete the initial configuration of the model; Step 2: Configure the compiler and compile the model into a.dll file, and import it into the multi-software simulation architecture after successful compilation; Step 3: Conduct model deployment and interface design in the multi-software simulation architecture, and monitor and test the model; Step 4: Use the parallel control system to test the actual control unit or control algorithm.

5. The active distribution network parallel control system with an energy storage configuration according to claim 4, characterized in that In step 3, based on a real-time simulation platform, a human-machine interaction interface is designed. This human-machine interaction interface can change the switch state during the simulation, set the active and reactive output powers of the load and distributed energy, and at the same time, with the help of the high computing power of the real-time simulator, it can feedback the node voltage change and the power of the distributed energy in real time, so as to modify the relevant control parameters or the state of the control unit in real time during the simulation, conduct real-time control of the system, and feedback the changes in the state of the distribution network before and after the parameter modification in real time.

6. The active distribution network parallel control system with energy storage configuration according to claim 4, wherein In step 4, during the operation of the parallel control system, the system operation parameters are modified, and the state changes of the system before and after are monitored in real time to achieve real-time simulation and real-time control of the system.

7. The active distribution network parallel control system with an energy storage configuration according to claim 6, characterized in that, In the active distribution network parallel control system with energy storage configuration, the multi-objective optimization configuration of the distribution network energy storage system and the decision-making of the optimal charge and discharge power strategy based on the Archimedes optimization algorithm are applied. The implementation method of this strategy decision-making includes the following steps: Step 1: Build a distribution network and energy storage system model with distributed energy access; Step 2: Propose a specific multi-objective evaluation function; Step 3: Apply the Archimedes optimization algorithm to optimize and solve the multi-objective model; Step 4: Output the optimal control decision information, and the control decision includes the decision information on the location and capacity determination of the energy storage system and the optimal charge and discharge power strategy information of the energy storage system.

8. The active distribution network parallel control system with an energy storage configuration according to claim 7, characterized in that, The specific process of building a distribution network and energy storage system model with distributed energy access in step 1 includes the following steps: Step (1-1): Build a distribution network system operation model, mainly including the fluctuation deviation index of the distribution network node voltage and the line loss rate during the operation of the distribution network; Step (1-2): Build an energy storage system configuration model, mainly including: the cost required for the entire life cycle operation of the energy storage, the output model of the energy storage system accessing the high-penetration distribution network; Step (1-3): Build a high-penetration distribution network operation model, an energy storage system access distribution network configuration model, and constraint conditions. The constraint conditions specifically include: (1) the node voltage constraint and line current constraint of the high-penetration distribution network with distributed energy access, (2) the power balance constraint, (3) the power flow constraint, (4) the energy storage battery capacity and energy multiple constraint, (5) the energy storage battery SOC state constraint, (6) the distributed energy output constraint.

9. The active distribution network parallel control system with an energy storage configuration according to claim 7, wherein The specific process of proposing a specific multi-objective evaluation function in step 2 includes: Step (2-1): Define a penalty objective function for the system security of the distribution network nodes; Step (2-2): Set a penalty objective function for the operation loss of the distribution network; Step (2-3): Consider the position of the energy storage system in the distribution network and set the evaluation objective function.

10. The active distribution network parallel control system with an energy storage configuration according to claim 7, wherein, The specific process of applying the Archimedes optimization algorithm to optimize and solve the multi-objective model in step 3 includes: Step (3-1): Initialize the parameters of the Archimedes optimization algorithm and the optimization objective problem, and use the normalization method to process the objectives in the multi-objective comprehensive optimization process; Step (3-2): Initialize the positions of all block individuals, including all decision variables, and initialize and randomly assign the object volume, density, and acceleration of each block running in water; Step (3-3) evaluates the initial population and selects the individual with the best fitness value to assign the best position, best density, best volume, and best acceleration, and updates the density and volume of each individual in the iteration. Step (3-4) defines the transfer operator and the density factor, and gradually transforms the algorithm from the search stage to the exploitation stage. Step (3-5) updates the acceleration of the wooden block individuals according to the mutual collisions between the wooden block individuals during the exploration stage; during the exploitation stage, there are no mutual collisions between the wooden blocks, and the "exploitation stage" formula is used to update the acceleration of the wooden block individuals. Step (3-6) normalizes the acceleration and calculates its percentage change. Step (3-7) updates the position of each individual and its affiliated attributes according to the transfer factor. Step (3-8) updates the fitness value of each individual using the objective function and records the individual with the best fitness value and its corresponding relevant attributes.

Images