Home V2H system load optimization control method based on master-slave game

Abstract

The invention is suitable for the technical field of smart power grid and home energy management crossing, and provides a home V2H system load optimization control method based on a master-slave game, and the method comprises the steps: constructing a master-slave game model of a home V2H system; a leader of the model is a household energy management system, two followers of the master-slave game model are an interruptible load cluster and a translatable load cluster, a target function of the leader aims at minimizing total household electricity cost in unit time, and a target function of the interruptible load cluster aims at maximizing earnings of the interruptible load cluster. The objective function of the translatable load cluster aims at maximizing the income of the translatable load cluster; solving Stackelberg balance of the master-slave game model to obtain an EV charging and discharging instruction and a load scheduling instruction; and the EV charging and discharging instruction and the load dispatching instruction are dynamically corrected, and the load is controlled based on the dynamic correction. According to the invention, the load optimization precision of the household V2H system can be improved.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of smart grid and home energy management, and particularly relates to a home V2H system load optimization control method based on a master-slave game. BACKGROUND

[0002] With the increase of the popularity of electric vehicles (EV) and the promotion of distributed photovoltaics, vehicle-to-home (V2H) technology has become a key support for realizing home energy two-way interaction. The V2H technology can realize energy exchange between the power grid and the home through the EV power battery, and further reduce the home electricity cost in cooperation with the solar photovoltaic (PV) module. The current load management of the home V2H system mostly adopts a centralized optimization strategy, which schedules all the loads, EVs and PVs as a unified whole, and does not fully consider the differences in the operating characteristics and decision autonomy of different types of loads.

[0003] In the prior art, some schemes introduce the game theory to realize load management, but mainly have the following defects: in the master-slave game model, the income function design between the home energy management system (HEMS) and the load is rough, and different target functions are not constructed for different response characteristics of interruptible loads (such as electric water heaters) and shiftable loads (such as washing machines), thereby causing low load optimization precision of the home V2H system. SUMMARY

[0004] The embodiment of the application provides a home V2H system load optimization control method based on a master-slave game, which can solve the problem of low load optimization precision of the home V2H system.

[0005] The embodiment of the application provides a home V2H system load optimization control method based on a master-slave game, which can solve the problem of low load optimization precision of the home V2H system.

[0006] A master-slave game model of the home V2H system is constructed; the leader of the master-slave game model is a home energy management system, the two followers of the master-slave game model are an interruptible load cluster and a shiftable load cluster, the target function of the home energy management system is a leader target function, the target function of the interruptible load cluster is an interruptible load income function, and the target function of the shiftable load cluster is a shiftable load income function; the leader target function aims to minimize the total electricity cost of the home per unit time, the interruptible load income function aims to maximize the income of the interruptible load cluster, and the shiftable load income function aims to maximize the income of the shiftable load cluster;

[0007] A Stackelberg equilibrium of the master-slave game model is solved by using a backward induction method, and EV charging and discharging instructions and load scheduling instructions are obtained;

[0008] When the control cycle arrives, the EV charging and discharging command and the load scheduling command are dynamically corrected to obtain the final EV charging and discharging command and the final load scheduling command. The home energy management system then controls the load of the home V2H system based on the final EV charging and discharging command and the final load scheduling command.

[0009] Optional, a home V2H system includes a home energy management system, an electric vehicle, a PV module, an interruptible load cluster, a shiftable load cluster, a bidirectional charger / discharger, and a grid interface.

[0010] The home energy management system connects to the power grid via a grid interface. It also connects to PV modules, interruptible load clusters, movable load clusters, bidirectional chargers, and electric vehicles. The output ports of the PV modules are connected to the power interfaces of the bidirectional chargers, interruptible load clusters, and movable load clusters, respectively. The charging interfaces of the bidirectional chargers are connected to the charging interfaces of the electric vehicles, interruptible load clusters, and movable load clusters, respectively.

[0011] Optionally, the interruptible load cluster includes multiple household appliances that can be temporarily de-energized or completely shut down during operation;

[0012] A movable load cluster includes multiple household appliances that need to maintain rated power operation during the performance of a task, and the operating time of the household appliances can be transferred within a specified time window.

[0013] Optionally, the leader's objective function is:

[0014] ;

[0015] in, express function, Indicates the scheduling period Total electricity cost within the area express Real-time grid electricity price express Real-time power exchange between the household and the power grid. express The cost of charging and discharging losses of electric vehicles at all times. express The cost of PV prediction error penalty at any given time. express The grid connection price of electricity generated by PV modules at any given time. express The actual power generation of the PV module at any given time.

[0016] Optional, Time-based PV prediction error penalty cost The calculation formula is:

[0017] ;

[0018] in, This represents the error penalty coefficient. express The predicted power generation of the PV module at any given time.

[0019] Optional, Cost of charging and discharging of electric vehicles The calculation formula is:

[0020] ;

[0021] in, express The charging and discharging power of electric vehicles at all times. This indicates the charging efficiency of electric vehicles. This indicates the discharge efficiency of an electric vehicle.

[0022] Optionally, the interruptible load revenue function is:

[0023] ;

[0024] in, express function, Indicates the scheduling period Benefits of interruptible load clusters This represents the unit power incentive price that a home energy management system sends to the interruptible load cluster. express The actual operating power of the load cluster can be interrupted at any time. This represents the interruption loss coefficient of an interruptible load cluster. This indicates the rated power of the interruptible load cluster.

[0025] Optionally, the revenue function for shiftable loads is:

[0026] ;

[0027] in, Indicates the scheduling period Benefits of internally movable load clusters This represents the unit power incentive price that a home energy management system sends to a cluster of movable loads. express The actual operating power of the load cluster can be shifted at any time. This represents the transfer cost coefficient for movable loads. This indicates the preset target power of the movable load cluster.

[0028] Optionally, the EV charging / discharging commands and load scheduling commands are dynamically revised to obtain the final EV charging / discharging commands and the final load scheduling commands, including:

[0029] Obtain the actual power generation of the PV module when the control cycle arrives;

[0030] The actual power generation and the predicted power generation of the PV module at the arrival of the control cycle are calculated. Based on the actual power generation and the predicted power generation, the PV prediction error is calculated. Based on the PV prediction error, the EV charging and discharging command and the load dispatching command are dynamically corrected to obtain the final EV charging and discharging command and the final load dispatching command.

[0031] Optionally, the EV charging / discharging command and load scheduling command are dynamically corrected based on the PV prediction error to obtain the final EV charging / discharging command and the final load scheduling command, including:

[0032] If the PV prediction error is less than or equal to the preset threshold, then the EV charging and discharging command and the load scheduling command will be directly used as the final EV charging and discharging command and the final load scheduling command.

[0033] If the PV prediction error is greater than a preset threshold, then the objective function of the leader is updated based on the PV prediction error. It then returns to the step of using backward induction to solve the Stackelberg equilibrium of the master-slave game model, obtaining EV charging / discharging instructions and load scheduling instructions, and uses the obtained EV charging / discharging instructions and load scheduling instructions as the final EV charging / discharging instructions and the final load scheduling instructions.

[0034] The above-mentioned solution in this application has the following beneficial effects:

[0035] In the embodiments of this application, a master-slave game model is constructed by separately constructing payoff functions for the characteristics of interruptible loads and movable loads. The interruptible load clusters and movable load clusters corresponding to these two payoff functions are considered as two independent followers, and the home energy management system (HEMS) with the objective of minimizing the total household electricity cost per unit time is considered as the leader. Finally, by solving the Stackelberg equilibrium of the master-slave game model, EV charging / discharging commands and load scheduling commands are obtained. Since this application constructs payoff functions for the characteristics of interruptible and movable loads respectively to participate in the game, it solves the problem of low optimization accuracy caused by homogeneous load responses in traditional game theory, effectively improving the load optimization accuracy of the household V2H system and making load scheduling more aligned with actual operational needs. Simultaneously, by deeply coupling the electric vehicle charging / discharging strategy with load scheduling through a master-slave game, it achieves coordinated control of energy flow and information flow with HEMS as the core, improving PV absorption rate.

[0036] Furthermore, by modifying the electric vehicle charging and discharging strategy and load scheduling when the control cycle arrives, this application is more adaptable to real-time changes in grid electricity prices and PV power compared to the prior art, thereby optimizing cost control.

[0037] Other beneficial effects of this application will be described in detail in the following detailed description section. Attached Figure Description

[0038] To more clearly illustrate the technical solutions in the embodiments of this application, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0039] Figure 1 An architectural diagram of a home V2H system provided in an embodiment of this application;

[0040] Figure 2 A flowchart of a load optimization control method for a home V2H system based on master-slave game theory provided in an embodiment of this application;

[0041] Figure 3 This is an overall flowchart of a home V2H system load optimization control method provided in an embodiment of this application. Detailed Implementation

[0042] In the following description, specific details such as particular system architectures and techniques are set forth for illustrative purposes and not for limitation, in order to provide a thorough understanding of the embodiments of this application. However, those skilled in the art will understand that this application may also be implemented in other embodiments without these specific details. In other instances, detailed descriptions of well-known systems, apparatuses, circuits, and methods have been omitted so as not to obscure the description of this application with unnecessary detail.

[0043] It should be understood that, when used in this application specification and the appended claims, the term "comprising" indicates the presence of the described features, integrals, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, integrals, steps, operations, elements, components and / or a collection thereof.

[0044] It should also be understood that the term “and / or” as used in this application specification and the appended claims means any combination of one or more of the associated listed items and all possible combinations, and includes such combinations.

[0045] As used in this application specification and the appended claims, the term "if" may be interpreted, depending on the context, as "when," "once," "in response to determination," or "in response to detection." Similarly, the phrase "if determined" or "if detected [the described condition or event]" may be interpreted, depending on the context, as meaning "once determined," "in response to determination," "once detected [the described condition or event]," or "in response to detection [the described condition or event]."

[0046] Furthermore, in the description of this application and the appended claims, the terms "first," "second," "third," etc., are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0047] References to "one embodiment" or "some embodiments" as described in this specification mean that one or more embodiments of this application include a specific feature, structure, or characteristic described in connection with that embodiment. Therefore, the phrases "in one embodiment," "in some embodiments," "in other embodiments," "in still other embodiments," etc., appearing in different parts of this specification do not necessarily refer to the same embodiment, but rather mean "one or more, but not all, embodiments," unless otherwise specifically emphasized. The terms "comprising," "including," "having," and variations thereof mean "including but not limited to," unless otherwise specifically emphasized.

[0048] To address the low load optimization accuracy of current residential V2H systems, this application provides a load optimization control method for residential V2H systems based on master-slave game theory. This method constructs payoff functions for interruptible and movable loads, respectively, and establishes a master-slave game model with the corresponding interruptible and movable load clusters as two independent followers, and the residential energy management system (HEMS) as the leader, aiming to minimize the total household electricity cost per unit time. Finally, by solving the Stackelberg equilibrium of the master-slave game model, EV charging / discharging commands and load scheduling commands are obtained. Since this application constructs payoff functions for interruptible and movable loads respectively to participate in the game, it solves the problem of low optimization accuracy caused by homogeneous load responses in traditional game theory, effectively improving the load optimization accuracy of residential V2H systems and making load scheduling more aligned with actual operational needs. Simultaneously, by deeply coupling the electric vehicle charging / discharging strategy with load scheduling through master-slave game theory, it achieves coordinated control of energy flow and information flow with HEMS as the core, improving PV absorption rate.

[0049] Furthermore, by modifying the electric vehicle charging and discharging strategy and load scheduling when the control cycle arrives, this application is more adaptable to real-time changes in grid electricity prices and PV power compared to the prior art, thereby optimizing cost control.

[0050] The home V2H system provided in this application will be described below with reference to specific embodiments.

[0051] like Figure 1 As shown, the home V2H system provided in this application embodiment includes a home energy management system (HEMS), an electric vehicle (EV), a solar photovoltaic (PV) power generation module (i.e., the photovoltaic power generation (PV) in the figure), an interruptible load cluster, a shiftable load cluster, a bidirectional charger and discharger, and a grid interface.

[0052] The home energy management system connects to the power grid via a grid interface. It also connects to PV modules, interruptible load clusters, movable load clusters, bidirectional chargers, and electric vehicles. The output ports of the PV modules are connected to the power interfaces of the bidirectional chargers, interruptible load clusters, and movable load clusters, respectively. The charging interfaces of the bidirectional chargers are connected to the charging interfaces of the electric vehicles, interruptible load clusters, and movable load clusters, respectively.

[0053] HEMS: It uses an STM32 series microcontroller as its core, integrates Ethernet, IoT (ZigBee) and Controller Area Network (CAN) bus interfaces, has real-time power dispatching and computing capabilities, supports grid-connected / off-grid mode switching, and realizes bidirectional power transmission control between EV and home power system through bidirectional charging and discharging piles. It connects to the power grid through the grid interface to obtain relevant information (such as grid time-of-use pricing) and power.

[0054] EV: Select a pure electric vehicle with a range of 500km, a power battery capacity of 75kWh, a charging and discharging power range of -15kW to 20kW (negative values ​​are for discharging, positive values ​​are for charging), and a state of charge (SOC) operating range of 20% to 90%.

[0055] PV module: It uses 20 300W monocrystalline silicon photovoltaic panels with a total power of 6kW. It is equipped with a power generation prediction module (prediction error ≤10%) and an anti-reverse current controller.

[0056] Interruptible load clusters include multiple household electrical appliances that can temporarily reduce power or be completely interrupted during operation. The interruption or power reduction does not affect the core performance of the equipment; only a minimum continuous operating time constraint must be met. Specific examples include electric water heaters (interruption compensation coefficient of 0.8 yuan / kWh), electric heaters, and other appliances. It is understood that interruptible load devices are pre-set with interruption compensation coefficients and minimum operating times. For example, the minimum operating time constraint for appliances such as electric water heaters and electric heaters is continuous operation for no less than 2 hours.

[0057] A movable load cluster includes multiple household appliances that need to maintain rated power operation during task execution, and their operating time can be transferred within a specified time window. The core requirement of these appliances is to complete their predetermined operational tasks. Specifically, this includes appliances such as washing machines (specified time window: 9:00-18:00) and dishwashers (specified time window: 18:00-21:00). It is understood that movable load devices are all pre-set with specified time windows and transfer cost coefficients. For example, the time window constraint for appliances such as washing machines and dishwashers can be 8:00-22:00.

[0058] Bidirectional charger / discharger: Supports V2H mode, AC input 220V, charging and discharging efficiencies are 0.92 and 0.9 respectively.

[0059] In some embodiments of this application, HEMS interacts with the power grid, PV modules, interruptible load clusters, transferable load clusters, electric vehicles, and bidirectional chargers / dischargers via a communication module to collect relevant information, including the short-term power generation prediction value and prediction error range of the PV modules, the initial SOC and charging / discharging power constraints of the EVs, the minimum operating time and interruption compensation coefficient of the interruptible loads, the time window (i.e., the aforementioned specified time window) and transfer cost coefficient of the transferable loads, as well as the time-of-use electricity price curve of the power grid, the interaction power between the household and the power grid, the grid-connected electricity price of the PV modules, the actual operating power, rated power, and preset target power of the interruptible load clusters and transferable load clusters, etc.

[0060] The following describes the load optimization control method for a home V2H system based on master-slave game theory provided in this application by way of specific embodiments.

[0061] like Figure 2 As shown in the embodiments of this application, the load optimization control method for a home V2H system based on master-slave game theory includes the following steps:

[0062] Step 21: Construct a master-slave game model for the home V2H system. The leader of this master-slave game model is the home energy management system, and the two followers are the interruptible load cluster and the movable load cluster. The objective function of the home energy management system is the leader's objective function, the objective function of the interruptible load cluster is the interruptible load revenue function, and the objective function of the movable load cluster is the movable load revenue function. The leader's objective function aims to minimize the total electricity cost of the household per unit time, the interruptible load revenue function aims to maximize the revenue of the interruptible load cluster, and the movable load revenue function aims to maximize the revenue of the movable load cluster.

[0063] In the aforementioned master-slave game model, HEMS is the leader, responsible for formulating EV charging and discharging strategies and load scheduling incentive signals; the interruptible load cluster and the shiftable load cluster are two independent followers, adjusting their own operating status according to the incentive signals.

[0064] The total electricity cost mentioned above includes grid purchase costs, EV charging and discharging losses, PV power generation revenue, and a PV prediction error penalty. Specifically, the leader's objective function is as follows:

[0065] ;

[0066] in, express function, Indicates the scheduling period Total electricity cost within the area express Real-time grid electricity price express The power exchange between the household and the power grid at any given time (positive value for purchasing electricity, negative value for selling electricity). express The cost of charging and discharging losses of electric vehicles at all times. express The cost of PV prediction error penalty at any given time. express The grid connection price of electricity generated by PV modules at any given time. express The actual power generation of the PV module at any given time. The scheduling cycle can be set according to the actual situation.

[0067] The above Time-based PV prediction error penalty cost The calculation formula is:

[0068] ;

[0069] in, This represents the error penalty coefficient (which can be preset based on experience or adjusted according to actual conditions, for example, set to 1.2-1.5). express The predicted power generation of the PV module at any given time.

[0070] Considering the differences in charge and discharge efficiency, the above Cost of charging and discharging of electric vehicles The calculation formula is:

[0071] ;

[0072] in, express The charging and discharging power of electric vehicles at any time (positive value for charging, negative value for discharging). This indicates the charging efficiency of electric vehicles. This indicates the discharge efficiency of an electric vehicle. and The value of is not less than 0.9.

[0073] In some embodiments of this application, the interruptible load cluster and the shiftable load cluster are two independent followers. The objective function of the interruptible load cluster is the interruptible load revenue function, and the objective function of the shiftable load cluster is the shiftable load revenue function.

[0074] The revenue of an interruptible load cluster is the difference between the incentive revenue and the interruption loss. Specifically, the revenue function for the interruptible load mentioned above is:

[0075] ;

[0076] in, express function, Indicates the scheduling period Benefits of interruptible load clusters This represents the unit power incentive price that a home energy management system sends to the interruptible load cluster. express The actual operating power of the load cluster can be interrupted at any time. This represents the interruption loss coefficient of an interruptible load cluster. This indicates the rated power of the interruptible load cluster.

[0077] Understandably, if the interruptible load cluster contains multiple devices, then the scheduling cycle for each device must be calculated separately according to the formula described above. The revenue within the period is calculated, and then the revenue of all devices is summed to obtain the interruptible load cluster's revenue during the scheduling cycle. The revenue generated within.

[0078] In some embodiments of this application, the revenue of a transferable load cluster is the difference between incentive revenue and transfer costs. Specifically, the transferable load revenue function is:

[0079] ;

[0080] in, Indicates the scheduling period Benefits of internally movable load clusters This represents the unit power incentive price that a home energy management system sends to a cluster of movable loads. express The actual operating power of the load cluster can be shifted at any time. This represents the transfer cost coefficient for movable loads. This indicates the preset target power of the movable load cluster.

[0081] The above This can be understood as the power that the system or user originally planned to operate at before the shiftable load participated in the game optimization (i.e., the preset target power). Its core is to serve as the "baseline power" after the shiftable load adjusts its operating state, which is used to calculate the cost of power transfer.

[0082] Step 22: Solve the Stackelberg equilibrium of the master-slave game model using backward induction to obtain EV charging and discharging commands and load scheduling commands.

[0083] The aforementioned EV charging and discharging instructions include: the EV charging period, the EV discharging period, the EV charging power, and the EV discharging power. These EV charging and discharging instructions not only meet the EV's own usage needs (such as maintaining SOC within a safe range) but also serve the overall energy optimization goals of the household. The aforementioned load dispatching instructions include: core control signals issued by the home energy management system to interruptible load clusters and shiftable load clusters. These core control signals mainly include: the controlled object (such as a washing machine or water heater), the running time, the operating mode, the incentive and constraint relationships, and the feedback status.

[0084] The operation mode is the specific operation instructions issued by HEMS based on the game optimization results: EV adjusts power according to the strategy of charging during off-peak hours and discharging during peak PV hours, interruptible loads can reduce power or be interrupted during peak hours (to meet the minimum running time), and loads can be shifted to transfer the running time to the peak PV period (to complete the task within the specified time window).

[0085] Incentive and constraint relationship: The incentive is based on the unit power incentive price sent by HEMS to the two types of loads, which directly affects the load response enthusiasm; the constraints include equipment constraints (such as minimum interruptible load operating time ≥ 2 hours, EVSOC range 20%-90%), optimization target constraints (serving the minimization of total electricity cost and maximization of PV absorption rate) and error accuracy constraints (PV prediction error ≤ 10%, load response deviation ≤ 5%).

[0086] Feedback: Feedback is the key to closed-loop control. HEMS collects data such as actual PV power, EV SOC, and load operating power in real time. If the PV prediction error exceeds the threshold, it triggers strategy correction and verifies the load response accuracy, forming a complete closed loop of "command issuance → execution → data feedback → correction" to adapt to real-time operating condition changes.

[0087] In some embodiments of this application, backward induction is a commonly used method for solving Stackelberg equilibrium in master-slave game models. The solution principle will not be elaborated upon here; only the solution process will be described as follows:

[0088] Step 22.1, fix the excitation signal and And EV charging and discharging strategies (which can be understood as charging and discharging time and power as initial values);

[0089] Step 22.2, maximize and Thus output and Maximum and Output and The core variable that influences total household electricity demand and, consequently, the leader's objective function. .

[0090] Step 22.3: Use the particle swarm optimization algorithm to solve for the optimal strategy of the leader, and obtain the EV charging and discharging command and load scheduling command.

[0091] When using the particle swarm optimization algorithm to solve for the optimal strategy, the traditional logic of the particle swarm optimization algorithm is adopted. In this process, each individual in the initial population includes EV charging / discharging instructions and load scheduling instructions, and the fitness function is the leader's objective function. Through particle swarm optimization iteration, the combination of decision variables that maximizes fitness (minimizes total electricity cost) is found (i.e., the EV charging / discharging instructions and load scheduling instructions obtained in step 22.3 above).

[0092] For example, in the execution of the above particle swarm optimization algorithm, the parameters are set as follows: population size 20-40, number of iterations 40-60, initial value of inertia weight 0.9, which is linearly decreased to 0.4 during the iteration process.

[0093] Step 23: When the control cycle arrives, the EV charging and discharging command and the load scheduling command are dynamically corrected to obtain the final EV charging and discharging command and the final load scheduling command. The home energy management system is then used to control the load of the home V2H system based on the final EV charging and discharging command and the final load scheduling command.

[0094] The aforementioned control cycle can be adjusted according to actual conditions, for example, set to 15 minutes. That is, after obtaining the EV charging / discharging commands and load scheduling commands by solving the Stackelberg equilibrium, the EV charging / discharging commands and load scheduling commands can be dynamically corrected at the end of each control cycle (e.g., every 15 minutes). It is understood that this correction process is generally performed before the next scheduling cycle arrives.

[0095] In some embodiments of this application, the specific implementation of dynamically modifying the EV charging / discharging command and load scheduling command in step 23 above to obtain the final EV charging / discharging command and final load scheduling command includes the following steps 23.1 and 23.2:

[0096] Step 23.1: Obtain the actual power generation of the PV module when the control cycle arrives.

[0097] Specifically, the actual power generation of the PV module at the arrival of the control cycle can be calculated by collecting the output voltage and output current of the PV module.

[0098] Step 23.2: Calculate the actual power generation and the predicted power generation of the PV module when the control cycle arrives, and calculate the PV prediction error based on the actual power generation and the predicted power generation. Based on the PV prediction error, dynamically correct the EV charging and discharging command and the load dispatching command to obtain the final EV charging and discharging command and the final load dispatching command.

[0099] In some embodiments of this application, the absolute error between the actual power generation and the predicted power generation can be used as the PV prediction error.

[0100] In some embodiments of this application, the specific implementation of dynamically correcting the EV charging / discharging command and load scheduling command based on the PV prediction error to obtain the final EV charging / discharging command and final load scheduling command is as follows:

[0101] If the PV prediction error is less than or equal to a preset threshold (e.g., 10%), then the EV charging / discharging command and the load scheduling command will be directly used as the final EV charging / discharging command and the final load scheduling command.

[0102] If the PV prediction error is greater than a preset threshold (e.g., 10%), then the objective function of the leader is updated based on this PV prediction error. Then return to step 22: use reverse induction to solve the Stackelberg equilibrium of the master-slave game model to obtain the EV charging and discharging instructions and load scheduling instructions, and use the obtained EV charging and discharging instructions and load scheduling instructions as the final EV charging and discharging instructions and the final load scheduling instructions.

[0103] Among them, the objective function for updating the leader based on the PV prediction error is as follows: This refers to: adjustment The error penalty coefficient in the calculation formula is used as the PV prediction error. .

[0104] The adjustment process for the error penalty coefficient is as follows:

[0105] The adjustment rule is as follows: When the PV prediction error is greater than the threshold (this value can be set according to the actual situation, such as 0.1), the error penalty coefficient is dynamically increased within the preset range (1.2-1.5) (following the principle that the larger the PV prediction error, the greater the increase in the error penalty coefficient. The core is to strengthen the constraint on the prediction deviation by increasing the penalty coefficient to avoid overspending on electricity costs) in order to ensure control stability and avoid load response disorder due to excessive penalty; when the PV prediction error is less than or equal to the threshold, the error penalty coefficient is maintained at the initial setting value.

[0106] Understandably, upon receiving the final EV charging / discharging command and the final load dispatching command, the home energy management system will control the load of the home V2H system based on these commands. That is, it controls the EVs based on the final EV charging / discharging command and controls the interruptible load clusters and shiftable load clusters based on the final load dispatching command.

[0107] In practical applications, HEMS controls the charging and discharging status of EVs through power electronic interfaces, sends operating commands to two types of loads (interruptible load clusters and movable load clusters) through the smart home bus, and collects power feedback signals from each unit in real time to form a closed-loop control.

[0108] For example, with a 24-hour scheduling cycle (T=96 15-minute time slots), the time-of-use electricity price is as follows: 0.3 yuan / kWh during off-peak hours (0:00-8:00), 0.6 yuan / kWh during normal hours (8:00-12:00, 18:00-22:00), and 0.9 yuan / kWh during peak hours (12:00-18:00, 22:00-24:00). The PV grid connection price is 0.45 yuan / kWh, and the error penalty coefficient is 1.3.

[0109] Initialization phase: HEMS collected PV predicted power to reach peak (5-6kW) between 10:00 and 15:00, EV initial SOC was 30%, interruptible loads were all in operation, and shiftable loads were initially scheduled to run between 12:00 and 13:00.

[0110] Follower response solution: Fixed HEMS initial incentive price It is 0.7 yuan / kWh. The price is 0.5 yuan / kWh. To maximize revenue, interruptible load clusters reduce power to 50% of the rated value during peak hours (12:00-18:00), while shiftable load clusters shift their operating time to the peak PV period (11:00-12:00).

[0111] Leader strategy optimization: Substitute the follower response into the objective function and solve it through particle swarm optimization (population size 30, number of iterations 50) to obtain the following: EVs are charged to 80% SOC at 20kW during off-peak hours (0:00-8:00), discharged to household loads at 15kW during peak PV hours (11:00-15:00), and charging is stopped during peak hours (12:00-18:00).

[0112] Dynamic correction: At 12:00, the actual PV power collected was 0.8kW lower than the predicted value (error 16% > 10%), triggering the adjustment of the penalty term. The error penalty coefficient was temporarily increased to 1.4. After resolving, HEMS increased the EV discharge power to 16kW to compensate for the PV power gap.

[0113] Execution and Feedback: HEMS controls the EV charging and discharging piles to execute commands via the CAN bus, sends power adjustment signals to the loads via ZigBee, and monitors the response deviation of each load in real time to be less than or equal to 5%, which meets the control accuracy requirements.

[0114] like Figure 3 As shown in the embodiments of this application, the load optimization control method for a home V2H system based on master-slave game theory constructs a master-slave game model with the Home Energy Management System (HEMS) as the leader and interruptible and portable load clusters as independent followers. The game equilibrium is solved by inverse induction and particle swarm optimization, and the strategy is dynamically adjusted every 15 minutes based on the actual PV power and EV SOC. Finally, the HEMS issues control commands to control the EVs and the interruptible and portable load clusters, and collects feedback signals.

[0115] This application solves the problems of low optimization accuracy and poor cost control in existing technologies by subdividing load characteristics and dynamic error compensation. It can reduce daily household electricity costs by more than 30% and increase PV self-consumption rate to over 90%, making it suitable for V2H systems in households with PV. Table 1 shows a comparison of various indicators between this embodiment and traditional non-game-theoretic optimization V2H systems.

[0116] Table 1

[0117]

[0118] In summary, the load optimization control method for a home V2H system based on master-slave game theory provided in this application has the following advantages:

[0119] 1) By adopting a differentiated follower payoff function design, target models are constructed for interruptible and shiftable loads respectively, which solves the problem of low optimization accuracy caused by homogeneous load response in traditional game theory, and makes load scheduling more in line with actual operation needs.

[0120] 2) A PV prediction error penalty mechanism is introduced. By dynamically adjusting the penalty coefficient, the prediction deviation is corrected in real time, which reduces the impact of PV randomness on electricity costs. Experimental results show that the fluctuation range of total electricity costs can be reduced by 15%-20%.

[0121] 3) By deeply coupling EV charging and discharging strategies with load scheduling through master-slave game theory, and taking HEMS as the core, the coordinated control of energy flow and information flow is realized, which improves PV absorption rate and can increase the self-consumption rate of household PV to more than 90% under typical lighting conditions.

[0122] 4) The 15-minute short-cycle dynamic correction strategy is adopted, which is more adaptable to the real-time changes in grid electricity price and PV power compared with traditional scheduling methods, and further optimizes the cost control effect.

[0123] The above description is the preferred embodiment of this application. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principles described in this application, and these improvements and modifications should also be considered within the scope of protection of this application.

Claims

1. A load optimization control method for a home V2H system based on a master-slave game, characterized in that, The application relates to a household V2H system, and a control method thereof. The master-slave game model is constructed; The leader of the master-slave game model is a household energy management system, two followers of the master-slave game model are an interruptible load cluster and a shiftable load cluster, a target function of the household energy management system is a leader target function, a target function of the interruptible load cluster is an interruptible load benefit function, and a target function of the shiftable load cluster is a shiftable load benefit function; the leader target function aims to minimize the total electricity cost of the household per unit time; the interruptible load benefit function aims to maximize the benefit of the interruptible load cluster; and the shiftable load benefit function aims to maximize the benefit of the shiftable load cluster. The Stackelberg equilibrium of the master-slave game model is solved by using a reverse induction method to obtain an EV charging and discharging instruction and a load scheduling instruction. When a control cycle arrives, the EV charging and discharging instruction and the load scheduling instruction are dynamically corrected to obtain a final EV charging and discharging instruction and a final load scheduling instruction, and the household energy management system controls the load of the household V2H system based on the final EV charging and discharging instruction and the final load scheduling instruction.

2. The home V2H system load optimization control method of claim 1, wherein, The household V2H system comprises a household energy management system, an electric vehicle, a PV module, an interruptible load cluster, a shiftable load cluster, a bidirectional charging and discharging machine and a power grid interface. The household energy management system is connected with the power grid through the power grid interface, and is connected with the PV module, the interruptible load cluster, the shiftable load cluster, the bidirectional charging and discharging machine and the electric vehicle respectively; the output port of the PV module is connected with the power interface of the bidirectional charging and discharging machine, the power interface of the interruptible load cluster and the power interface of the shiftable load cluster respectively; and the charging interface of the bidirectional charging and discharging machine is connected with the charging interface of the electric vehicle, the power interface of the interruptible load cluster and the power interface of the shiftable load cluster respectively. 3.The home V2H system load optimization control method of claim 2, wherein, The interruptible load cluster comprises a plurality of household electrical equipment which can temporarily reduce power or completely interrupt during operation. The shiftable load cluster comprises a plurality of household electrical equipment which needs to maintain rated power operation during task execution, and the operation time of the household electrical equipment can be shifted within a specified time window. 4.The home V2H system load optimization control method of claim 2, wherein, The leader target function is: ； wherein, represents a function, represents the total electricity cost within the scheduling period, represents the grid electricity price at time instant, represents the interaction power between the home and the grid at time instant, represents the charging / discharging loss cost of the electric vehicle at time instant, represents the PV prediction error penalty cost at time instant, represents the grid-connected electricity price of the PV module at time instant, represents the actual power generation of the PV module at time instant. 5.The home V2H system load optimization control method of claim 4, wherein, Instantaneous PV prediction error penalty cost The calculation formula is: ； wherein, denotes an error penalty coefficient, denotes the predicted power generation of the PV module at the instant 6.The home V2H system load optimization control method of claim 5, wherein, Cost of electric vehicle charging and discharging loss at time The calculation formula is: ； wherein, represents represents the charging power of the electric vehicle at the time point, represents the charging efficiency of the electric vehicle, represents the discharging efficiency of the electric vehicle. 7.The home V2H system load optimization control method of claim 6, wherein, The interruptible load benefit function is: ； wherein, represents function, represents the dispatching period the benefit of the interruptible load cluster, represents the unit power incentive price sent by the home energy management system to the interruptible load cluster, represents the actual operating power of the interruptible load cluster at the moment, represents the interruption loss coefficient of the interruptible load cluster, represents the rated power of the interruptible load cluster. 8.The home V2H system load optimization control method of claim 7, wherein, The shiftable load benefit function is: ； wherein, denotes a dispatching period denotes the benefit of the movable load cluster, denotes the unit power incentive price sent by the home energy management system to the movable load cluster, denotes denotes the actual operating power of the movable load cluster at the moment, denotes the transfer cost coefficient of the movable load, denotes the preset target power of the movable load cluster. 9.The home V2H system load optimization control method of claim 8, wherein, The dynamic correction of the EV charging and discharging instruction and the load scheduling instruction to obtain the final EV charging and discharging instruction and the final load scheduling instruction comprises: The actual power generation of the PV module when the control cycle arrives is obtained; The actual power generation and the predicted power generation of the PV module when the control cycle arrives are calculated, a PV prediction error is calculated based on the actual power generation and the predicted power generation, and the EV charging and discharging instruction and the load scheduling instruction are dynamically corrected based on the PV prediction error to obtain the final EV charging and discharging instruction and the final load scheduling instruction. 10.The home V2H system load optimization control method of claim 9, wherein, The EV charging and discharging instruction and the load scheduling instruction are dynamically corrected based on the PV prediction error to obtain a final EV charging and discharging instruction and a final load scheduling instruction, including: If the PV prediction error is less than or equal to a preset threshold, the EV charging and discharging instruction and the load scheduling instruction are directly taken as the final EV charging and discharging instruction and the final load scheduling instruction. If the PV prediction error is greater than a preset threshold, the target function of the leader is updated based on the PV prediction error And return to execute the step of solving the master-slave game model using the backward induction method to obtain the EV charging and discharging instruction and the load scheduling instruction, and the obtained EV charging and discharging instruction and load scheduling instruction are taken as the final EV charging and discharging instruction and the final load scheduling instruction.

Images