A multi-mesh distributed shared scheduling method and device

Abstract

This application discloses a distributed shared scheduling method and apparatus for multiple microgrids, relating to the field of grid charging, and addressing issues such as existing energy management schemes not considering range anxiety, privacy protection, and ESS degradation costs. The method constructs a target state-of-charge curve based on charging behavior data and a preset anxiety mapping relationship, generating a range anxiety penalty cost. Combining the dynamic power boundary of energy storage and degradation costs, it constructs an optimization model aimed at minimizing the local overall operating cost. By receiving the global average tie-line power and dual variables from the coordination center, the model is solved to obtain the expected tie-line exchange power, which is then uploaded and iteratively updated until convergence. This invention achieves distributed optimized scheduling that balances user range anxiety and energy storage lifetime, effectively ensuring the economy and stability of microgrid operation.

Description

Technical Field

[0001] This application relates to the field of power grid charging, and in particular to a method and apparatus for distributed shared scheduling of multiple microgrids. Background Technology

[0002] As the penetration rate of electric vehicles (EVs) in microgrids continues to increase, large-scale unregulated EV charging is causing a surge in local loads, posing a serious challenge to the grid's supply and demand balance. Stationary energy storage systems (ESS) are considered as two complementary flexible resources, and their coordinated scheduling is key to the stable operation of the system.

[0003] However, existing energy management strategies face significant limitations when dealing with complex user behaviors and multi-microgrid interconnected scenarios. First, traditional scheduling strategies typically assume ideal EV response compliance, ignoring the actual impact of user "range anxiety." In practice, vehicle owners tend to charge conservatively, limiting the EV's adjustment capabilities in terms of both time and power. This oversight overestimates the system's flexibility and severely shifts the burden of mitigating load fluctuations to stationary energy storage systems, leading to accelerated lifespan degradation due to high-frequency charging and discharging. Second, constrained by conservative charging behavior and uneven energy storage distribution within individual microgrids, individual microgrids lack sufficient capacity to manage load fluctuations. Furthermore, as the scale of EVs grows, traditional centralized scheduling faces the curse of dimensionality and computational challenges in handling large-scale discrete variables, and requires microgrids to share fine-grained operational data, resulting in serious privacy risks.

[0004] Therefore, providing a scheduling method that considers range anxiety and ESS constraints to achieve distributed energy sharing and privacy protection among multiple microgrids is a technical problem that urgently needs to be solved by those in the field. Summary of the Invention

[0005] The purpose of this application is to provide a distributed shared scheduling method and device for multiple microgrids, which solves the problems of existing energy management schemes that do not consider range anxiety, privacy protection, and ESS degradation costs.

[0006] To address the aforementioned technical problems, this application provides a multi-micronet distributed shared scheduling method and apparatus, applied to each micronet side, comprising: To obtain operational data of this micronetwork and charging behavior data of each electric vehicle within this micronetwork; Based on the charging behavior data, a target state of charge curve is constructed using a preset range anxiety mapping relationship; Based on the target state of charge curve and the actual state of charge of the electric vehicle battery, a deviation variable characterizing the user's range anxiety is generated, and the range anxiety penalty cost is obtained based on the deviation variable. Based on the aforementioned operational data, the dynamic power boundary and degradation cost of the fixed energy storage system within this microgrid are determined; With the goal of minimizing local overall operating costs, a local optimized scheduling model is constructed that includes the dynamic power boundary; wherein, the local overall operating costs include: net electricity purchase cost, range anxiety penalty cost, and degradation cost; Receive global average tie-line power and dual variable signals from the coordination center; Based on the global average tie-line power and the dual variable, the local optimal scheduling model is solved to obtain the expected tie-line switching power of this micronetwork. The expected switching power of the tie line is sent to the coordination center, so that the coordination center updates the global average tie line power and the dual variable according to the expected switching power of the tie line, and returns to the step of receiving the global average tie line power and dual variable signal sent by the coordination center, until the convergence condition is met. After receiving the stop iteration command from the coordination center, the local decision variables of this micronetwork are output as scheduling commands.

[0007] Optionally, in the above multi-microgrid distributed shared scheduling method, the charging behavior data includes the grid entry time and grid exit time of each electric vehicle; Based on the charging behavior data, a target state of charge curve is constructed using a preset range anxiety mapping relationship, including: The normalized time progress parameters are determined based on the network entry time and network exit time. The normalized time schedule parameter is as follows: ; In the formula, This represents the normalized time schedule parameter; Indicates the current scheduling period; This represents the time when the i-th electric vehicle joins the network; This represents the time when the i-th electric vehicle leaves the network. ; i represents the electric vehicle number; A normalized anxiety function representing the mileage anxiety mapping relationship is established based on the normalized time progress parameters; The normalized anxiety function is: ; In the formula, This represents the normalized anxiety function; Indicates the anxiety and urgency level, and 0; The target curve of the state of charge is established based on the normalized anxiety function; The target curve for the state of charge is represented as follows: ; In the formula, This represents the target curve of the state of charge; This indicates the initial state of charge of the electric vehicle; This indicates the target state of charge when the electric vehicle leaves.

[0008] Optionally, in the above multi-microgrid distributed shared scheduling method, the generation of deviation variables characterizing user range anxiety based on the target state of charge curve and the actual state of charge of the electric vehicle battery includes: Construct the actual state of charge of electric vehicle batteries based on their physical parameters; The actual state of charge is: ; In the formula, Indicates the first The car is The actual state of charge at any given moment; This represents the charging power of the i-th electric vehicle during time period t; Indicates the charging efficiency of electric vehicles; Indicates the scheduling time step; Indicates the rated capacity of the battery; The difference between the actual state of charge and the target state of charge curve is used as the deviation variable; Correspondingly, the mileage anxiety penalty cost is obtained based on the aforementioned deviation variable, including: The deviation variables are relaxed by applying linear inequality constraints to obtain the relaxed deviation variables. The linear inequality constraint is as follows: ; ; In the formula, Indicates the deviation variable; This represents the target curve of the state of charge; Indicates the actual state of charge; A secondary penalty is applied to the relaxed deviation variable to obtain the mileage anxiety penalty cost; The cost of the range anxiety penalty is as follows: ; In the formula, This represents the cost of the range anxiety penalty; Indicates the conversion weighting coefficient; This indicates the number of electric vehicles.

[0009] Optionally, in the above-mentioned multi-microgrid distributed shared scheduling method, determining the dynamic power boundary and degradation cost of the fixed energy storage system within the microgrid based on the operational data includes: The charging and discharging power of the stationary energy storage system is constrained by a state-of-charge-dependent dynamic charging and discharging power boundary, which includes a dynamic charging power boundary and a dynamic discharging power boundary. The dynamic boundary of the charging power is: ; The dynamic boundary of the discharge power is: ; In the formula, This indicates the charging power; This indicates the discharge power; Indicates the rated maximum power; Indicates the dynamic constraint coefficient; Indicates the percentage of maximum state of charge; Indicates the minimum percentage of the state of charge; Indicates the rated capacity of the energy storage system; This indicates the energy state in the previous time period; The degradation cost is determined based on the equivalent depreciation cost factor per unit power throughput of the stationary energy storage system, the charging power, the discharging power, and the first formula. The first formula is: ; In the formula, This represents the degradation cost; The equivalent depreciation cost factor representing the power throughput per unit; This indicates the total number of time periods in the scheduling cycle.

[0010] Optionally, in the above multi-micronet distributed shared scheduling method, with the objective of minimizing local overall operating costs, a local optimized scheduling model including the dynamic power boundary is constructed, including: The net electricity purchase cost is determined based on the basic tiered electricity purchase volume, the penalty tiered electricity purchase volume, and the electricity sold to the main grid; An objective function for the local integrated operating cost is established based on the net electricity purchase cost, the range anxiety penalty cost, and the degradation cost. The objective function is: ; In the formula, This indicates the net cost of electricity purchase; This indicates the cost of mileage anxiety; Indicates the cost of degradation; The objective function is denoted as .

[0011] Optionally, the above-mentioned multi-micronet distributed shared scheduling method also includes: A power balance constraint is established based on the premise that the sum of all power supplied within this microgrid equals the sum of all power consumed. The power balance constraint condition is: ; In the formula, This indicates the net switching power with the main network; Indicates photovoltaic power generation capacity; This indicates the amount of solar power curtailment; Indicates the discharge power of a stationary energy storage system; Indicates the base load power; Indicates the charging power of a stationary energy storage system; Indicates the switching power of the tie line; This represents the charging power of the i-th electric vehicle; Establish boundary constraints for electric vehicle charging power and electric vehicle charge state based on electric vehicle charging behavior data. Based on the operational data of the stationary energy storage system in this microgrid, establish the energy state boundary constraints and energy state transition constraints of the stationary energy storage system.

[0012] Optionally, in the above-mentioned distributed shared scheduling method for multiple micronets, the step of solving the local optimal scheduling model based on the global average tie-line power and the dual variable to obtain the expected tie-line switching power of the micronet includes: An augmented Lagrangian function is constructed based on the objective function, the global average tie-line power, and the dual variables. The augmented Lagrange function is: ; In the formula, Indicates the first The augmented Lagrangian function of a microgrid; This represents the local overall operating cost of this micronetwork; Let represent the vector of local decision variables for the m-th microgrid; Indicates the ADMM penalty parameter; This represents the tie-line switching power variable to be optimized. It represents the residual power determined by the global average tie-line power and the tie-line exchange power of the m-th microgrid in the k-th iteration; Let represent the dual variable of the k-th iteration; k represents the iteration number. With the goal of minimizing the augmented Lagrange function, the updated expected switching power of the tie lines in this micronetwork is obtained by solving the problem.

[0013] To address the aforementioned technical problems, this application also provides a multi-micronet distributed shared scheduling method, applied to the coordination center side, comprising: Send the initial global average tie-line power and dual variables to each microgrid; The expected switching power of the tie lines is obtained by solving the local optimization scheduling model of each microgrid based on the global average tie line power and the dual variable. The global average tie-line power is updated based on the received tie-line expected switching power of each micronet. The dual variable is updated based on the updated global average tie-line power. The original residual is determined based on the expected switching power of the contact lines of each microgrid, and it is determined whether the original residual is less than a preset accuracy threshold. If the original residual is less than the preset accuracy threshold, a stop iteration command is sent to each microgrid. If the original residual is greater than or equal to the preset accuracy threshold, then based on the updated global average tie-line power and the updated dual variable signal as the current round signal, the global average tie-line power and dual variable are sent to each microgrid based on the current round signal, and the process of solving the local optimization scheduling model of each microgrid to obtain the expected tie-line exchange power is returned.

[0014] Optionally, in the above multi-micronet distributed shared scheduling method, updating the global average tie-line power based on the received tie-line expected switching power of each micronet includes: The updated global average tie-line power is obtained by taking the arithmetic mean of the expected switching power of each microgrid tie-line. Correspondingly, updating the dual variable based on the updated global average tie-line power includes: Add the current dual variable to the updated global average tie-line power to obtain the updated dual variable; Correspondingly, determining the original residual based on the received expected switching power of the tie lines of each micronet includes: The norm of the sum of the expected switching power of each microgrid's tie lines is obtained based on the expected switching power of each microgrid's tie lines received. The norm is used as the original residual.

[0015] To address the aforementioned technical problems, this application also provides a multi-micronet distributed shared scheduling device, applied to each micronet side, comprising: The acquisition module is used to acquire the operation data of this micronetwork and the charging behavior data of each electric vehicle within this micronetwork; The target charging curve construction module is used to construct a target state of charge curve based on the charging behavior data and using a preset range anxiety mapping relationship. The range anxiety quantification module is used to generate a deviation variable characterizing the user's range anxiety based on the target state of charge curve and the actual state of charge of the electric vehicle battery, and to obtain the range anxiety penalty cost based on the deviation variable. The energy storage degradation cost quantification module is used to determine the dynamic power boundary and degradation cost of the fixed energy storage system within this microgrid based on the operating data. The local cost quantification module is used to construct a local optimized scheduling model that includes the dynamic power boundary with the goal of minimizing the local comprehensive operating cost; wherein, the local comprehensive operating cost includes: net electricity purchase cost, range anxiety penalty cost, and degradation cost; The receiving module is used to receive the global average tie-line power and dual variable signals sent by the coordination center; The optimization module is used to solve the local optimization scheduling model based on the global average tie-line power and the dual variable to obtain the expected tie-line switching power of this micronetwork. The update module is used to send the expected switching power of the tie line to the coordination center, so that the coordination center updates the global average tie line power and the dual variable according to the expected switching power of the tie line, and returns to the step of receiving the global average tie line power and dual variable signal sent by the coordination center, until the convergence condition is met. The output module is used to receive the stop iteration command issued by the coordination center and output the local decision variables of this micronetwork as scheduling instructions.

[0016] The multi-microgrid distributed shared scheduling method provided in this application is applied to each microgrid side. By acquiring the microgrid's operational data and electric vehicle charging behavior data, and using a preset range anxiety mapping relationship to construct a target state of charge curve, it reflects users' psychological expectations of fast charging in the early stages and slow charging in the later stages, avoiding the problem of overestimating the flexibility of electric vehicles due to ignoring range anxiety. By generating a deviation variable representing users' range anxiety based on the target state of charge curve and the actual state of charge of the electric vehicle battery, and obtaining the range anxiety penalty cost based on this deviation variable, users' psychological expectations are transformed into quantifiable economic costs. Therefore, in subsequent optimizations, user charging needs will be proactively met, and user satisfaction will be directly incorporated into the optimization objectives, reducing deviations caused by users' lack of cooperation in scheduling. By determining the dynamic power boundary of the fixed energy storage system based on operational data, the charging and discharging power is automatically limited when the energy storage system's state of charge approaches the upper or lower limit, preventing overcharging and over-discharging damage and extending the service life of the fixed energy storage system. At the same time, the degradation cost of the fixed energy storage system is determined and quantified into an economic cost to suppress high-frequency ineffective charging and discharging of the energy storage system, reducing maintenance and replacement costs. By constructing a local optimal scheduling model with dynamic power boundaries to minimize local overall operating costs—including net electricity purchase costs, range anxiety penalty costs, and degradation costs—an intrinsic balance among the three objectives is achieved, avoiding over-regulation that might result from single-objective optimization. Each microgrid receives the global average tie-line power and dual variable signals from the coordination center and solves the local optimal scheduling model based on these global variables to obtain its expected tie-line exchange power. Each microgrid only needs to report its expected tie-line exchange power to the coordination center, without including any internal operating data such as photovoltaic output, load curves, electric vehicle status, or energy storage status, thus achieving data privacy protection among multiple microgrids. Furthermore, since only scalar signals are transmitted in each iteration, the communication bandwidth requirement is significantly reduced. After receiving the stop-iteration command from the coordination center, the local decision variables of the microgrid are output as scheduling instructions, ensuring that the tie-line exchange plans of each microgrid are matched and guaranteeing that the scheduling plan meets the global scheduling requirements.

[0017] In addition, this application also provides a multi-micronet distributed shared scheduling method and a multi-micronet distributed shared scheduling device applied to the coordination center side, which correspond to the above-mentioned multi-micronet distributed shared scheduling method applied to each micronet side and have the same effect. Attached Figure Description

[0018] To more clearly illustrate the embodiments of this application, the accompanying drawings used in the embodiments 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.

[0019] Figure 1 A flowchart illustrating a multi-micronet distributed shared scheduling method applied to various micronet sides, provided in this application embodiment; Figure 2(a) is a distribution diagram of PV output and foundation load of MG1 at different time periods provided in an embodiment of this application; Figure 2(b) is a distribution diagram of PV output and foundation load of MG2 at different time periods provided in an embodiment of this application; Figure 2(c) is a distribution diagram of PV output and foundation load of MG3 at different time periods provided in an embodiment of this application; Figure 2(d) is a distribution diagram of the spatiotemporal distribution of EVs of MG1 provided in an embodiment of this application; Figure 2(e) is a distribution diagram of the spatiotemporal distribution of EVs of MG2 provided in an embodiment of this application; Figure 2(f) is a distribution diagram of the spatiotemporal distribution of EVs of MG3 provided in an embodiment of this application; Figure 3 This is a flowchart illustrating a distributed shared scheduling method for multiple micronets applied to the coordination center side, as described in this application embodiment. Figure 4 This is a schematic diagram of a multi-micronet distributed shared scheduling device applied to each micronet side, provided as an embodiment of this application. Detailed Implementation

[0020] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the protection scope of this application.

[0021] The core of this application is to provide a method and apparatus for distributed shared scheduling of multiple micronets.

[0022] To enable those skilled in the art to better understand the present application, the present application will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0023] This application provides a distributed shared scheduling method for multiple micronets, applicable to each micronet side, such as... Figure 1 As shown, it includes: S11: Obtain the operation data of this micronetwork and the charging behavior data of each electric vehicle within this micronetwork; S12: Based on charging behavior data, construct a target state of charge curve using a preset range anxiety mapping relationship; S13: Generate a deviation variable characterizing the user's range anxiety based on the target state of charge curve and the actual state of charge of the electric vehicle battery, and obtain the range anxiety penalty cost based on the deviation variable. S14: Determine the dynamic power boundary and degradation cost of the stationary energy storage system within this microgrid based on operational data; S15: To minimize the local comprehensive operating cost, a local optimal scheduling model with dynamic power boundaries is constructed; the local comprehensive operating cost includes: net electricity purchase cost, range anxiety penalty cost, and degradation cost. S16: Receive the global average tie-line power and dual variable signals issued by the coordination center; S17: Based on the global average tie-line power and dual variables, solve the local optimal scheduling model to obtain the expected tie-line switching power of this microgrid. S18: Send the expected switching power of the tie line to the coordination center so that the coordination center updates the global average tie line power and dual variable according to the expected switching power of the tie line, and return to the step of receiving the global average tie line power and dual variable signal sent by the coordination center until the convergence condition is met. S19: After receiving the stop iteration instruction from the coordination center, output the local decision variables of this micronetwork as scheduling instructions.

[0024] This embodiment is primarily implemented in a multi-microgrid system with high electric vehicle penetration, such as a system consisting of residential microgrids, industrial microgrids, and office microgrids interconnected in a chain structure via tie lines (the number of interconnected microgrids is at least two). Each microgrid integrates a distributed photovoltaic power generation array, baseload, stationary energy storage system, and a group of electric vehicles. The capacity of the stationary energy storage system can vary among the microgrids; for example, residential microgrids may be equipped with large-capacity stationary energy storage systems, industrial microgrids with medium-capacity stationary energy storage systems, and office microgrids may not be equipped with stationary energy storage systems.

[0025] Figure 2(a) is a distribution diagram of PV output and base load of MG1 at different time periods provided in an embodiment of this application; Figure 2(b) is a distribution diagram of PV output and base load of MG2 at different time periods provided in an embodiment of this application; Figure 2(c) is a distribution diagram of PV output and base load of MG3 at different time periods provided in an embodiment of this application; Figure 2(d) is a distribution diagram of the spatiotemporal distribution of EVs of MG1 provided in an embodiment of this application; Figure 2(e) is a distribution diagram of the spatiotemporal distribution of EVs of MG2 provided in an embodiment of this application; Figure 2(f) is a distribution diagram of the spatiotemporal distribution of EVs of MG3 provided in an embodiment of this application; As shown in Figures 2(a)-2(f), Residential parking mode (MG1): EVs mainly arrive and charge in the evening and leave the next morning; Industrial shift mode (MG2): EV arrival and departure behavior is aligned with the factory shift schedule; Office commuting mode (MG3): EVs arrive during the morning working hours and leave after get off work in the evening.

[0026] First, each microgrid needs to acquire its own operational data and the charging behavior data of each electric vehicle within the microgrid. Operational data may include information such as distributed photovoltaic power generation, base load power, and the state of charge (SOC) of stationary energy storage systems. Charging behavior data may include the grid connection time, grid disconnection time, initial SOC, and target SOC upon departure of each electric vehicle. This data provides the basic input for subsequent scheduling decisions.

[0027] Based on the acquired charging behavior data, each microgrid constructs a target state of charge curve using a pre-defined range anxiety mapping relationship. It should be noted that traditional scheduling strategies typically assume that electric vehicle users completely obey scheduling instructions, using only the final state of charge when the vehicle leaves as a constraint, ignoring the user's psychological need to reach a safe charge level as quickly as possible during parking. This embodiment quantifies this range anxiety by constructing a target state of charge curve. Specifically, this curve starts at the initial state of charge of the electric vehicle and ends at the target state of charge upon departure, exhibiting a faster rate of increase in the early stage than in the later stage. This reflects the user's behavioral pattern of wanting to obtain more charge in the initial charging stage to alleviate range anxiety, and then gradually balancing to the target charge level later.

[0028] Based on the target state of charge (SOC) curve and the actual SOC of the electric vehicle battery, each microgrid generates a deviation variable characterizing user range anxiety, and derives a range anxiety penalty cost based on this deviation variable. The actual SOC must adhere to the battery's inherent physical characteristics, determined by factors such as charging power, charging efficiency, and battery rated capacity. The deviation variable represents the positive difference between the target SOC curve and the actual SOC, indicating the degree to which the actual charge is lower than the user's expected charge. When the actual SOC reaches or exceeds the expected value, the deviation variable is zero. A secondary penalty is applied to this deviation variable to obtain the range anxiety penalty cost. Since this penalty cost uses a quadratic function, the convexity of the optimization process is ensured, while small deviations are tolerated and large deviations are strictly penalized, guiding the scheduling strategy to prioritize meeting users' charging expectations. Smaller deviations indicate that the scheduling strategy better meets user needs, thereby increasing users' responsiveness to microgrid scheduling.

[0029] Each microgrid also determines the dynamic power boundary and degradation cost of its stationary energy storage system based on operational data. Stationary energy storage systems face physical limitations during charging and discharging. When the battery is near full charge, charging power needs to be limited to prevent overcharging; when the battery is near depletion, discharging power needs to be limited to prevent over-discharging. This characteristic is known as constant current / constant voltage charging in battery management. This embodiment represents this characteristic through a dynamic power boundary, allowing the upper limit of charging and discharging power to dynamically adjust with changes in the state of charge (SOC). The closer the SOC is to the boundary, the lower the allowable charging and discharging power. Furthermore, stationary energy storage systems experience lifespan degradation during frequent charging and discharging. This embodiment quantifies the charging and discharging throughput as degradation cost. Specifically, the sum of the charging and discharging power for each scheduling period, multiplied by a unit depreciation factor, is summed over each scheduling period to obtain the total degradation cost. By introducing this cost term into the optimization objective, the system's excessive reliance on and ineffective use of stationary energy storage systems can be suppressed, balancing the contradiction between adjustment flexibility and equipment lifespan economy.

[0030] Each microgrid aims to minimize its overall local operating cost by constructing a local optimization scheduling model that includes dynamic power boundaries. The overall local operating cost comprises three parts: net electricity purchase cost, range anxiety penalty cost, and degradation cost. Net electricity purchase cost covers the basic electricity exchange between the microgrid and the main grid, excess penalty electricity, electricity sales revenue, and photovoltaic curtailment penalty. Range anxiety penalty cost, as mentioned above, reflects the cost of unmet user expectations. Degradation cost reflects the lifespan loss of stationary energy storage systems due to charging and discharging. These three cost items are physically interdependent: prioritizing electric vehicle charging reduces range anxiety penalty cost but may increase net electricity purchase cost; over-reliance on stationary energy storage systems increases degradation cost; reducing electricity purchases from the main grid reduces net electricity purchase cost but may require more use of local resources. Through joint optimization of these three factors, the system can automatically find a balance between electricity purchase economics, user satisfaction, and equipment lifespan.

[0031] In the distributed solution process, each microgrid receives the global average tie-line power and dual variable signals from the coordination center. Based on these global variables, each microgrid solves its local optimal scheduling model to obtain its expected tie-line exchange power. This expected tie-line exchange power is then sent to the coordination center. The coordination center updates the global average tie-line power and dual variables based on the received expected tie-line exchange power from each microgrid. This process is repeated iteratively until convergence is met. It's important to note that in this distributed architecture, each microgrid reports only the scalar signal of expected tie-line exchange power, excluding any internal operational data such as photovoltaic output, load curves, electric vehicle status, or stationary energy storage system status. This achieves data privacy protection among multiple stakeholders and significantly reduces communication bandwidth requirements. Local solutions in each microgrid can be performed in parallel, with computational complexity increasing linearly with the number of microgrids rather than exponentially, overcoming the dimensionality curse faced by centralized scheduling. Furthermore, the coordination center only aggregates power signals; even in the event of a brief loss of connection, each microgrid can continue operating based on its local data and the previous round of global variables, avoiding the risk of system paralysis due to a single point of failure.

[0032] Once the coordination center determines that the convergence condition is met, it issues a stop-iteration command to each microgrid. Upon receiving this command, each microgrid outputs its local decision variables as scheduling instructions. These local decision variables may include the charging power of each electric vehicle in each scheduling period, the charging and discharging power of the stationary energy storage system in each scheduling period, the interaction power between the microgrid and the main grid, the exchange power between the microgrid and the tie line, the amount of abandoned photovoltaic power, and the range anxiety deviation variable of each electric vehicle in each scheduling period. These instructions can be directly sent to execution devices such as charging piles and energy storage converters.

[0033] Based on the above principles, the multi-microgrid distributed shared scheduling method provided in this embodiment acquires the microgrid's operational data and electric vehicle charging behavior data, and constructs a target state of charge curve using a preset range anxiety mapping relationship to reflect users' psychological expectations of fast charging in the early stages and slow charging in the later stages, avoiding the problem of overestimating the flexibility of electric vehicles due to ignoring range anxiety. By generating a deviation variable representing users' range anxiety based on the target state of charge curve and the actual state of charge of the electric vehicle battery, and obtaining the range anxiety penalty cost based on this deviation variable, users' psychological expectations are transformed into quantifiable economic costs. Therefore, in subsequent optimizations, user charging needs will be proactively met, and user satisfaction will be directly incorporated into the optimization objectives, reducing deviations caused by users' lack of cooperation in scheduling. By determining the dynamic power boundary of the fixed energy storage system based on operational data, the charging and discharging power is automatically limited when the energy storage system's state of charge approaches the upper or lower limit, preventing overcharging and over-discharging damage and extending the service life of the fixed energy storage system. At the same time, the degradation cost of the fixed energy storage system is determined and quantified into economic costs to suppress high-frequency ineffective charging and discharging of the energy storage system, reducing maintenance and replacement costs. By constructing a local optimal scheduling model with dynamic power boundaries to minimize local overall operating costs—including net electricity purchase costs, range anxiety penalty costs, and degradation costs—an intrinsic balance among the three objectives is achieved, avoiding over-regulation that might result from single-objective optimization. Each microgrid receives the global average tie-line power and dual variable signals from the coordination center and solves the local optimal scheduling model based on these global variables to obtain its expected tie-line exchange power. Each microgrid only needs to report its expected tie-line exchange power to the coordination center, without including any internal operating data such as photovoltaic output, load curves, electric vehicle status, or energy storage status, thus achieving data privacy protection among multiple microgrids. Furthermore, since only scalar signals are transmitted in each iteration, the communication bandwidth requirement is significantly reduced. After receiving the stop-iteration command from the coordination center, the local decision variables of the microgrid are output as scheduling instructions, ensuring that the tie-line exchange plans of each microgrid are matched and guaranteeing that the scheduling plan meets the global scheduling requirements.

[0034] Based on the above embodiments, this embodiment further defines the specific implementation method of the range anxiety mapping relationship, and the charging behavior data includes the grid access time and grid disconnection time of each electric vehicle; Based on charging behavior data, a target state of charge curve is constructed using a pre-defined range anxiety mapping relationship, including: The normalized time progress parameters are determined based on the network entry time and network exit time. The normalized time schedule parameter is as follows: ; In the formula, This represents the normalized time schedule parameter; Indicates the current scheduling period; This represents the time when the i-th electric vehicle joins the network; This represents the time when the i-th electric vehicle leaves the network. ; i represents the electric vehicle number; A normalized anxiety function representing the mapping relationship of mileage anxiety is established based on normalized time schedule parameters; The normalized anxiety function is: ; In the formula, Represents the normalized anxiety function; Indicates the anxiety and urgency level, and 0; Establish a target curve for the state of charge based on a normalized anxiety function; The target curve for the state of charge is represented as follows: ; In the formula, Represents the target curve of the state of charge; This indicates the initial state of charge of the electric vehicle; This indicates the target state of charge when the electric vehicle leaves.

[0035] The normalized time schedule parameter maps the charging windows of different electric vehicles to a uniform interval of zero to one, where A value of 0 indicates that the vehicle has just been connected to the power grid. A value of 1 indicates that the vehicle is about to leave the grid. This allows subsequent expected state-of-charge trajectories to be compared and calculated on the same time reference, regardless of differences in charging window length.

[0036] A normalized anxiety function is established based on normalized time schedule parameters to represent the mapping relationship of mileage anxiety. This represents the anxiety and urgency coefficient, which is greater than zero and is used to control the rate of ascent and curvature of the target state of charge curve. The faster the target state of charge curve rises in the early stages of the charging window, the higher the user's range anxiety level; when... As the value approaches 0, the curve approaches the linear charging expectation.

[0037] When the anxiety urgency coefficient is positive, the normalized anxiety function rises rapidly in the early stage of the charging window and then slows down in the later stage, causing the target state of charge curve to exhibit a rapid increase in the early stage and a slow approach to the target state of charge in the later stage. This characteristic is consistent with the user's psychological expectations: in the early stage of the charging window, the user hopes that the battery level will increase rapidly to alleviate range anxiety; in the later stage of the charging window, the user's sense of urgency regarding the charging speed decreases, and they can accept a slower charging rate.

[0038] Continuing to establish the target state of charge curve based on the normalized anxiety function, when the vehicle is first connected, the normalized time progress parameter... The normalized anxiety function is 0. A value of 0 indicates that the target state of charge curve equals the initial state of charge, meaning the user expects to immediately reach the target charge level. When the vehicle is about to leave, the normalized time progress parameter... The value is 1, at which point the target state of charge curve equals the target state of charge. In reality, during the charging process, the target state of charge curve gradually rises from the initial state of charge to the target state of charge, and under the influence of the anxiety and urgency coefficient, it shows a trend of rapid rise in the early stage and gradual slowdown in the later stage. This guides the system to prioritize meeting the user's strong energy replenishment needs in the early stage of the charging window, and smoothly approach the target state of charge in the later stage.

[0039] Based on the above embodiments, this embodiment further defines the generation method of the range anxiety deviation variable and the calculation method of the range anxiety penalty cost. The deviation variable characterizing the user's range anxiety is generated based on the target state of charge curve and the actual state of charge of the electric vehicle battery, including: Construct the actual state of charge of electric vehicle batteries based on their physical parameters; The actual state of charge is: ; In the formula, Indicates the first The car is The actual state of charge at any given moment; This represents the charging power of the i-th electric vehicle during time period t; Indicates the charging efficiency of electric vehicles; Indicates the scheduling time step; Indicates the rated capacity of the battery; The difference between the actual state of charge and the target state of charge curve is used as the deviation variable. Correspondingly, the cost of mileage anxiety penalty is derived based on the deviation variable, including: The deviation variables are relaxed by applying linear inequality constraints to obtain the relaxed deviation variables. The linear inequality constraints are as follows: ; ; In the formula, Indicates the deviation variable; Represents the target curve of the state of charge; Indicates the actual state of charge; By applying a second penalty to the relaxed bias variable, the mileage anxiety penalty cost is obtained. The cost of mileage anxiety penalty is as follows: ; In the formula, This indicates the cost of mileage anxiety; Indicates the conversion weighting coefficient; This indicates the number of electric vehicles.

[0040] First, the actual state of charge (SPC) of the electric vehicle battery is constructed based on its physical parameters. The SPC is obtained by adding the charging power multiplied by the charging efficiency and the scheduling time step to the previous SPC, and then dividing by the battery's rated capacity. This reflects the quantitative relationship between energy input and changes in SPC.

[0041] In addition, the actual charging power is subject to the rated power of the charger, and the actual charging power must not exceed the rated power of the charger. ; This indicates the rated power of the charger.

[0042] The actual state of charge must be maintained within the physical limits: .

[0043] The difference between the actual state of charge (SOC) and the target SOC curve is used as the deviation variable. It's important to note that when the actual SOC exceeds the target curve, the user actually receives more power than expected, and there is no range anxiety. Only when the actual SOC falls below the target curve does a level of anxiety requiring quantification arise. Therefore, the deviation variable should only reflect positive differences; negative differences should be zeroed out. ; This represents the range anxiety bias variable. The target state-of-charge curve, while ensuring the travel needs of individual EV users, avoids the coordination center's blind overestimation of EV flexibility resources, achieving a good match between theoretical optimization and actual engineering operation.

[0044] To facilitate the embedding of this nonlinear relationship into a continuous quadratic programming optimization model, this embodiment employs linear inequality constraints to relax the deviation variables. Specifically, as shown in the linear inequality constraint, the deviation variable is greater than or equal to zero, and the deviation variable is greater than or equal to the target state of charge curve minus the actual state of charge. Since the deviation variable in the optimization objective appears in the form of a quadratic penalty and its weight coefficient is positive, the optimization solution will automatically make the deviation variable take the minimum value that satisfies the constraints, that is, exactly equal to the positive part of the target state of charge curve minus the actual state of charge, while the negative part is zero. Transforming it into a linear inequality constraint allows the model to be solved using an efficient quadratic programming algorithm.

[0045] Applying a secondary penalty to the relaxed deviation variable yields the range anxiety penalty cost. This secondary penalty ensures the convexity of the optimization problem, facilitating solution and guaranteeing global optimality. The secondary penalty exhibits higher tolerance for small deviations and stronger penalties for large deviations, aligning with actual user psychology—minor battery depletion is acceptable, but severe battery shortages trigger intense anxiety. By adjusting the conversion weight coefficients, the system's trade-off between user satisfaction and cost-effectiveness can be optimized.

[0046] Furthermore, this embodiment further defines the specific method for determining the dynamic power boundary and degradation cost of the stationary energy storage system. Determining the dynamic power boundary and degradation cost of the stationary energy storage system within this microgrid based on operational data includes: The charging and discharging power of a fixed energy storage system is constrained by a dynamic charging and discharging power boundary that depends on the state of charge. The dynamic charging and discharging power boundary includes the dynamic charging power boundary and the dynamic discharging power boundary. The dynamic boundary of charging power is: ; The dynamic boundary of discharge power is: ; In the formula, Indicates charging power; Indicates discharge power; Indicates the rated maximum power; This represents the dynamic constraint coefficient, β>0; Indicates the percentage of maximum state of charge; Indicates the minimum percentage of the state of charge; Indicates the rated capacity; This indicates the energy state in the previous time period; The degradation cost is determined based on the equivalent depreciation cost factor per unit power throughput of the stationary energy storage system, the charging power, the discharging power, and the first formula. The first formula is: ; In the formula, This represents the degradation cost of stationary energy storage systems; The equivalent depreciation cost factor representing the power throughput per unit; This indicates the total number of time periods in the scheduling cycle.

[0047] In this embodiment, the stationary energy storage system is subject to dynamic power boundary constraints dependent on its state of charge during charging and discharging, including dynamic boundaries for charging and discharging power. The energy state evolution equation for the stationary energy storage system is established as follows: ; in, for Energy status of stationary energy storage systems at the end of the time period; and These represent the charging power and discharging power during time period t, respectively. The charge-discharge cycle efficiency of a fixed energy storage system; The scheduling interval.

[0048] The energy state of stationary energy storage systems must meet physical capacity boundary constraints: ; in, For the rated capacity of a fixed energy storage system, and These represent the minimum and maximum percentage of state of charge, respectively.

[0049] In the dynamic boundary of charging power This represents a dynamic limit coefficient, a positive number, used to control the rate of decrease in charging or discharging power when the state of charge (SBC) of a stationary energy storage system approaches its upper or lower limit. This mechanism simulates the constant-current, constant-voltage charging characteristics of a battery: in the constant-current phase, the battery charges and discharges at its rated maximum power; when the SBC approaches its upper or lower limit, it enters the constant-voltage phase, where the power needs to decrease linearly to prevent overcharging or over-discharging. By introducing the dynamic power boundary, the charging and discharging plan generated by the scheduling model naturally conforms to the physical limitations of the battery, avoiding the risk of equipment damage caused by forced scheduling.

[0050] The degradation cost reflects the impact of charging and discharging frequency and intensity on battery life. Since the degradation cost in the optimization objective is positively correlated with charging and discharging power, the system will proactively reduce unnecessary charging and discharging operations when formulating scheduling plans. For example, it will avoid frequent charging and discharging of stationary energy storage systems when electricity price fluctuations are small, or prioritize mutual assistance through microgrid interconnects rather than calling local stationary energy storage systems to meet load demands. This achieves economic protection for stationary energy storage systems.

[0051] According to the above embodiments, specifically, with the goal of minimizing the local overall operating cost, a local optimized scheduling model including dynamic power boundaries is constructed, including: The net electricity purchase cost is determined based on the basic tiered electricity purchase volume, the penalty tiered electricity purchase volume, and the electricity sold to the main grid. Establish an objective function for local integrated operating costs based on net electricity purchase cost, range anxiety penalty cost, and degradation cost; The objective function is: ; In the formula, This indicates the net cost of electricity purchase; This indicates the cost of mileage anxiety; Indicates the cost of degradation; This represents the objective function.

[0052] First, the net electricity purchase cost is determined based on the basic tiered electricity purchase volume, the penalty tiered electricity purchase volume, and the volume sold to the main grid. The calculation method for the net electricity purchase cost is as follows: ; in, for Electricity price at any given time; Indicates the basic tiered electricity purchase volume; This indicates a tiered penalty system for purchasing electricity. This indicates that electricity has been sold to the main network; This indicates the penalty multiplier for exceeding the electricity purchase quota; Indicates the electricity sales multiplier; and This represents the amount of photovoltaic curtailment and the photovoltaic curtailment penalty coefficient.

[0053] The relationship between net electricity purchase cost and grid interconnection power is as follows: ; Boundary constraints for basic tiered electricity purchase: ; Boundary constraints on solar power curtailment: ; in, Indicates the basic upper limit for electricity purchase; This indicates the power output of photovoltaic power generation.

[0054] When the purchased electricity volume is within the basic purchase limit, a lower basic electricity price applies; when the purchased electricity volume exceeds the basic purchase limit, a higher penalty price applies to the excess portion. Furthermore, when the microgrid has surplus electricity, it can sell it to the main grid to generate revenue. Therefore, the net electricity purchase cost needs to incorporate the basic purchase price, penalty purchase price, and electricity sales together. The introduction of photovoltaic curtailment penalties encourages the system to prioritize the consumption of renewable energy generation, reducing energy waste.

[0055] Then, a target function for the local integrated operating cost is established based on the net electricity purchase cost, mileage anxiety penalty cost, and degradation cost. F represents the target function for the local integrated operating cost. The target function unifies the three dimensions of electricity purchase economy, user satisfaction, and equipment lifespan into a single objective function. During the solution process, the conflicts between the three are automatically weighed to find the overall optimal scheduling scheme.

[0056] Preferably, the local integrated operating cost also includes: the operation and maintenance cost of distributed photovoltaic power within the microgrid and the fuel cost of conventional units within the microgrid.

[0057] Furthermore, it also includes: A power balance constraint is established based on the premise that the sum of all power supplied within this microgrid equals the sum of all power consumed. The power balance constraint is: ; In the formula, This indicates the net switching power with the main network; Indicates photovoltaic power generation capacity; This indicates the amount of solar power curtailment; Indicates the discharge power of a stationary energy storage system; Indicates the base load power; Indicates the charging power of a stationary energy storage system; Indicates the switching power of the tie line; This represents the charging power of the i-th electric vehicle; Establish boundary constraints for electric vehicle charging power and electric vehicle charge state based on electric vehicle charging behavior data. Based on the operational data of the stationary energy storage system in this microgrid, establish the energy state boundary constraints and energy state transition constraints of the stationary energy storage system.

[0058] Power supply includes net exchange power with the main grid, effective photovoltaic power after deducting curtailed photovoltaic power generation, and discharge power from stationary energy storage systems. Power consumption includes base load power, charging power of stationary energy storage systems, tie-line exchange power, and the sum of charging power for all electric vehicles. Power balance constraints are the fundamental physical constraints of power system dispatch, ensuring energy conservation at all times.

[0059] Based on electric vehicle charging behavior data, boundary constraints for electric vehicle charging power and electric vehicle state of charge are established. The charging power boundary constraints are as follows: As shown, the charging power of each electric vehicle is required to be between zero and the rated power of the charger. The state-of-charge boundary constraints are as follows: As shown, each electric vehicle is required to have an actual state of charge between 0 and 1. These constraints ensure the safety of the electric vehicle charging process and the physical limits of the equipment hardware.

[0060] Based on the operational data of the stationary energy storage system in this microgrid, establish the energy state boundary constraints and energy state transition constraints for the stationary energy storage system. The energy state boundary constraints are as follows: As shown, the energy state of a stationary energy storage system is required to be between the rated capacity multiplied by the minimum state of charge percentage and the rated capacity multiplied by the maximum state of charge percentage. The energy state transition constraint is as follows: As shown, this transfer equation reflects the energy accumulation and release process of a stationary energy storage system. Through complete modeling of the above constraints, the local optimal scheduling model achieves physical feasibility, and the generated scheduling plan can be safely executed on actual equipment.

[0061] In one specific embodiment, based on the global average tie-line power and dual variables, the local optimal scheduling model is solved to obtain the expected tie-line switching power of the micronetwork, including: An augmented Lagrangian function is constructed based on the objective function, global average tie-line power, and dual variables. The augmented Lagrangian function is: ; In the formula, Indicates the first The augmented Lagrangian function of a microgrid; This represents the local overall operating cost of this micronetwork; Let represent the vector of local decision variables for the m-th microgrid; Indicates the ADMM penalty parameter; This represents the tie-line switching power variable to be optimized. It represents the residual power determined by the global average tie-line power and the tie-line exchange power of the m-th microgrid in the k-th iteration; Let represent the dual variable of the k-th iteration; k represents the iteration number. The expected exchange power of the tie lines in this micronetwork is obtained by minimizing the augmented Lagrangian function.

[0062] This represents the local comprehensive operating cost of this micronetwork, which is F in the objective function.

[0063] The residual power is calculated as follows: ; In the formula, This represents the average power of all microgrid interconnects in the k-th iteration; This represents the tie-line exchange power of the m-th microgrid in the k-th iteration.

[0064] It should be noted that the core of the Alternating Direction Multiplier Method (ADMM) is to decompose the global tie-line power balance constraint into local subproblems of each microgrid. The global tie-line power balance constraint is used to ensure that the sum of the tie-line exchange power of each microgrid approaches zero, meaning that the total output power among the microgrids matches the total input power. Each microgrid, while minimizing its own overall operating cost, adjusts its local tie-line exchange power based on the global average tie-line power and dual variables issued by the coordination center, to gradually reduce the global tie-line power imbalance. The local decision variable vector includes the electric vehicle charging power, stationary energy storage system charging and discharging power, interaction power with the main grid, tie-line exchange power, and photovoltaic curtailment power for each scheduling period.

[0065] The globally average tie-line power is used as a coordination reference for iterative updates of each microgrid, and does not imply that the tie-line exchange power of each microgrid must ultimately be equal. By introducing a quadratic penalty term and iterative updates of the dual variable, the sum of the expected tie-line exchange power of each microgrid gradually approaches zero, thus satisfying the power-sharing balance constraint among multiple microgrids. The updated expected tie-line exchange power of this microgrid is obtained by minimizing the augmented Lagrangian function. This solution process is a convex quadratic programming problem, which can be efficiently solved using mature numerical optimization algorithms. Each microgrid executes this solution process in parallel without interfering with each other.

[0066] The convergence condition is determined by whether the original residual is less than the preset accuracy threshold.

[0067] The original residuals are: ; In the formula, This represents the original residual in the (k+1)th iteration. Indicates the total number of microgrids; This represents the expected switching power of the tie lines in the (k+1)th iteration of the m-th microgrid. This indicates an assignment operation. This represents the Euclidean norm of a vector.

[0068] When the original residual is less than the preset accuracy threshold, each microgrid stops iterating and outputs its local decision variables.

[0069] This application also provides a multi-micronet distributed shared scheduling method, applied to the coordination center side, such as... Figure 3 As shown, it includes: S21: Send the initial global average tie-line power and dual variables to each microgrid; S22: Receive the expected switching power of each microgrid by solving the local optimization scheduling model of each microgrid based on the global average tie-line power and dual variables; S23: Update the global average tie-line power based on the received tie-line expected switching power of each micronet; S24: Update the dual variables based on the updated global average tie-line power; S25: Determine the original residual based on the expected switching power of each microgrid's interconnection line, and determine whether the original residual is less than the preset accuracy threshold. S26: If the original residual is less than the preset accuracy threshold, a stop iteration command is sent to each microgrid; S27: If the original residual is greater than or equal to the preset accuracy threshold, then based on the updated global average tie-line power and the updated dual variable signal as the current round signal, send the global average tie-line power and dual variable to each microgrid based on the current round signal, and return to solve the local optimization scheduling model of each microgrid to obtain the expected tie-line exchange power.

[0070] The coordination center first sends the initial global average tie-line power and dual variables to each microgrid. Generally, the initial value of the global average tie-line power can be set to zero, and the initial value of the dual variables can also be set to zero. This indicates that no consensus has been reached regarding the tie-line power before the iteration begins.

[0071] The coordination center receives the expected tie-line switching power obtained by each microgrid after solving its local optimal scheduling model based on the global average tie-line power and dual variables. Based on the received expected tie-line switching power from each microgrid, the coordination center updates the global average tie-line power. Then, based on the updated global average tie-line power, the coordination center updates the dual variables.

[0072] The coordination center determines the initial residual based on the expected switching power of each microgrid's tie lines received. That is, the formula... When the sum of the expected exchange power of the interconnect lines of each microgrid approaches zero, it indicates that a power balance consensus has been reached among the microgrids, meaning that the total planned power sent out equals the total planned power received. The coordination center determines whether the original residual is less than a preset accuracy threshold. The preset accuracy threshold can be pre-set according to the scheduling accuracy requirements, for example, it can be set to 2% or lower.

[0073] If the original residual is less than the preset accuracy threshold, it indicates that the convergence condition has been met, and the coordination center issues a stop iteration command to each micronet. If the original residual is greater than or equal to the preset accuracy threshold, it indicates that consensus has not yet been reached, and the coordination center uses the updated global average tie-line power and the updated dual variable signal as the current round signal, returning to the step of sending the initialized global average tie-line power and dual variable to each micronet to continue the next round of iteration. Through the above process, the coordination center only needs to process the expected tie-line exchange power reported by each micronet, without accessing any micronet's internal operating data, thus protecting the data privacy of multiple stakeholders while achieving global coordination.

[0074] Specifically, based on the received expected switching power of each micronet's tie lines, the global average tie line power is updated, including: The updated global average tie-line power is obtained by taking the arithmetic mean of the expected switching power of each microgrid tie-line. Correspondingly, based on the updated global average tie-line power, the dual variables are updated, including: Add the current dual variable to the updated global average tie-line power to obtain the updated dual variable; Correspondingly, the original residuals are determined based on the expected switching power of each microgrid's tie lines, including: The norm of the sum of the expected switching power of each microgrid's tie lines is obtained based on the expected switching power of each microgrid's tie lines received. The norm is used as the original residual.

[0075] The global average tie-line power is updated based on the received tie-line expected switching power of each micronet: ; In the formula, This represents the global average tie-line power in the (k+1)th iteration. This averaging operation is the core of the shared alternating direction multiplier method, which aggregates the independent decisions of each micronet into a consensus reference point.

[0076] Update the dual variable based on the updated global average tie-line power: ; In the formula, Let represent the dual variable in the (k+1)th iteration. The dual variable plays a role in accumulating historical biases in the alternating direction multiplier method. By accumulating the dual variable, the coordination center can record the average trend of tie-line power in each iteration and feed this information back to each microgrid, guiding local decisions to gradually adjust towards a globally consistent direction.

[0077] The norm of the sum of the expected switch power of each microgrid's tie lines is obtained based on the received expected switch power of each microgrid's tie lines, and this norm is used as the original residual. Ideally, when global convergence occurs, the sum of the expected switch power of each microgrid's tie lines should be zero, indicating that the planned power transmission and reception of all microgrids are matched. The original residual reflects the degree of inconsistency between the plans of each microgrid in the current iteration round. As the iteration progresses, the original residual gradually decreases. When it decreases to below a preset accuracy threshold, it can be considered that each microgrid has reached an acceptable consensus on the tie line switch power.

[0078] The multi-micronet distributed shared scheduling method has been described in detail in the above embodiments. This application also provides embodiments corresponding to the multi-micronet distributed shared scheduling device. It should be noted that this application describes the embodiments of the device part from two perspectives: one is based on functional modules, and the other is based on hardware.

[0079] From the perspective of functional modules Figure 4 A structural diagram of a multi-micronet distributed shared scheduling device provided in this application embodiment is shown below. Figure 4 As shown, a multi-micronet distributed shared scheduling device, applied to each micronet side, includes: The acquisition module 31 is used to acquire the operation data of this micronetwork and the charging behavior data of each electric vehicle in this micronetwork; The target charging curve construction module 32 is used to construct a target state of charge curve based on charging behavior data and using a preset range anxiety mapping relationship. The range anxiety quantification module 33 is used to generate deviation variables that characterize the user's range anxiety based on the target state of charge curve and the actual state of charge of the electric vehicle battery, and to obtain the range anxiety penalty cost based on the deviation variables. The energy storage degradation cost quantification module 34 is used to determine the dynamic power boundary and degradation cost of the fixed energy storage system in this microgrid based on the operating data; The local cost quantification module 35 is used to construct a local optimized scheduling model with dynamic power boundaries with the goal of minimizing the local comprehensive operating cost; wherein, the local comprehensive operating cost includes: net electricity purchase cost, range anxiety penalty cost, and degradation cost; Receiver module 36 is used to receive the global average tie-line power and dual variable signals sent by the coordination center; Optimization module 37 is used to solve the local optimal scheduling model based on the global average tie-line power and dual variables to obtain the expected tie-line switching power of this microgrid. The update module 38 is used to send the expected switching power of the tie line to the coordination center, so that the coordination center updates the global average tie line power and dual variable according to the expected switching power of the tie line, and returns to the step of receiving the global average tie line power and dual variable signal sent by the coordination center until the convergence condition is met. Output module 39 is used to receive the stop iteration command issued by the coordination center and output the local decision variables of this micronetwork as scheduling instructions.

[0080] The above modules can be deployed in the local controllers of each micronet, and the local controllers of each micronet are connected to the coordination center through a communication network.

[0081] Since the embodiments of the apparatus and the embodiments of the method correspond to each other, please refer to the description of the embodiments of the method for the embodiments of the apparatus, which will not be repeated here.

[0082] The multi-micronet distributed shared scheduling method and apparatus provided in this application have been described in detail above. The various embodiments in the specification are described in a progressive manner, with each embodiment focusing on its differences from other embodiments. Similar or identical parts between embodiments can be referred to interchangeably. For the apparatus disclosed in the embodiments, since it corresponds to the method disclosed in the embodiments, the description is relatively simple; relevant parts can be referred to in the method section. It should be noted that those skilled in the art can make several improvements and modifications to this application without departing from the principles of this application, and these improvements and modifications also fall within the protection scope of this application.

[0083] It should also be noted that, in this specification, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes the element.

Claims

1. A distributed shared scheduling method for multiple micronets, characterized in that, Applied to various microgrid sides, including: To obtain operational data of this micronetwork and charging behavior data of each electric vehicle within this micronetwork; Based on the charging behavior data, a target state of charge curve is constructed using a preset range anxiety mapping relationship; Based on the target state of charge curve and the actual state of charge of the electric vehicle battery, a deviation variable characterizing the user's range anxiety is generated, and the range anxiety penalty cost is obtained based on the deviation variable. Based on the aforementioned operational data, the dynamic power boundary and degradation cost of the fixed energy storage system within this microgrid are determined; With the goal of minimizing local overall operating costs, a local optimized scheduling model is constructed that includes the dynamic power boundary; wherein, the local overall operating costs include: net electricity purchase cost, range anxiety penalty cost, and degradation cost; Receive global average tie-line power and dual variable signals from the coordination center; Based on the global average tie-line power and the dual variable, the local optimal scheduling model is solved to obtain the expected tie-line switching power of this micronetwork. The expected switching power of the tie line is sent to the coordination center, so that the coordination center updates the global average tie line power and the dual variable according to the expected switching power of the tie line, and returns to the step of receiving the global average tie line power and dual variable signal sent by the coordination center, until the convergence condition is met. After receiving the stop iteration command from the coordination center, the local decision variables of this micronetwork are output as scheduling commands.

2. The multi-micronet distributed shared scheduling method according to claim 1, characterized in that, The charging behavior data includes the charging time and the charging time of each electric vehicle; Based on the charging behavior data, a target state of charge curve is constructed using a preset range anxiety mapping relationship, including: The normalized time progress parameters are determined based on the network entry time and network exit time. The normalized time schedule parameter is as follows: ; In the formula, This represents the normalized time schedule parameter; Indicates the current scheduling period; This represents the time when the i-th electric vehicle joins the network; This represents the time when the i-th electric vehicle leaves the network. , where i represents the electric vehicle number; A normalized anxiety function representing the mileage anxiety mapping relationship is established based on the normalized time progress parameters; The normalized anxiety function is: ; In the formula, This represents the normalized anxiety function; Indicates the anxiety and urgency level, and 0; The target curve of the state of charge is established based on the normalized anxiety function; The target curve for the state of charge is represented as follows: ; In the formula, This represents the target curve of the state of charge; This indicates the initial state of charge of the electric vehicle; This indicates the target state of charge when the electric vehicle leaves.

3. The multi-micronet distributed shared scheduling method according to claim 2, characterized in that, Based on the target state of charge curve and the actual state of charge of the electric vehicle battery, bias variables characterizing user range anxiety are generated, including: Construct the actual state of charge of electric vehicle batteries based on their physical parameters; The actual state of charge is: ; In the formula, Indicates the first The car is The actual state of charge at any given moment; This represents the charging power of the i-th electric vehicle during time period t; Indicates the charging efficiency of electric vehicles; Indicates the scheduling time step; Indicates the battery's rated capacity; The difference between the actual state of charge and the target state of charge curve is used as the deviation variable; Correspondingly, the mileage anxiety penalty cost is obtained based on the aforementioned deviation variable, including: The deviation variables are relaxed by applying linear inequality constraints to obtain the relaxed deviation variables. The linear inequality constraint is as follows: ； ； In the formula, Indicates the deviation variable; This represents the target curve of the state of charge; Indicates the actual state of charge; A secondary penalty is applied to the relaxed deviation variable to obtain the mileage anxiety penalty cost; The cost of the range anxiety penalty is as follows: ; In the formula, This represents the cost of the range anxiety penalty; Indicates the conversion weighting coefficient; This indicates the number of electric vehicles.

4. The multi-micronet distributed shared scheduling method according to claim 3, characterized in that, Based on the aforementioned operational data, the dynamic power boundary and degradation cost of the stationary energy storage system within this microgrid are determined, including: The charging and discharging power of the stationary energy storage system is constrained by a state-of-charge-dependent dynamic charging and discharging power boundary, which includes a dynamic charging power boundary and a dynamic discharging power boundary. The dynamic boundary of the charging power is: ; The dynamic boundary of the discharge power is: ; In the formula, This indicates the charging power; This indicates the discharge power; Indicates the rated maximum power; Indicates the dynamic constraint coefficient; Indicates the percentage of maximum state of charge; Indicates the minimum percentage of the state of charge; Indicates the rated capacity of the energy storage system; This indicates the energy state in the previous time period; The degradation cost is determined based on the equivalent depreciation cost factor per unit power throughput of the stationary energy storage system, the charging power, the discharging power, and the first formula. The first formula is: ; In the formula, This represents the degradation cost; The equivalent depreciation cost factor representing the power throughput per unit; This represents the total number of time periods in the scheduling cycle, t=1,2,…,T.

5. The multi-micronet distributed shared scheduling method according to claim 4, characterized in that, To minimize the overall local operating cost, a local optimization scheduling model incorporating the dynamic power boundary is constructed, including: The net electricity purchase cost is determined based on the basic tiered electricity purchase volume, the penalty tiered electricity purchase volume, and the electricity sold to the main grid; An objective function for the local integrated operating cost is established based on the net electricity purchase cost, the range anxiety penalty cost, and the degradation cost. The objective function is: ; In the formula, This indicates the net cost of electricity purchase; This indicates the cost of mileage anxiety; Indicates the cost of degradation; Let represent the objective function.

6. The multi-micronet distributed shared scheduling method according to claim 5, characterized in that, Also includes: A power balance constraint is established based on the premise that the sum of all power supplied within this microgrid equals the sum of all power consumed. The power balance constraint condition is: ; In the formula, This indicates the net switching power with the main network; Indicates photovoltaic power generation capacity; This indicates the amount of solar power curtailment; Indicates the discharge power of a stationary energy storage system; Indicates the base load power; Indicates the charging power of a stationary energy storage system; Indicates the switching power of the tie line; This represents the charging power of the i-th electric vehicle; Establish boundary constraints for electric vehicle charging power and electric vehicle charge state based on electric vehicle charging behavior data. Based on the operational data of the stationary energy storage system in this microgrid, establish the energy state boundary constraints and energy state transition constraints of the stationary energy storage system.

7. The multi-micronet distributed shared scheduling method according to claim 5, characterized in that, The step of solving the local optimal scheduling model based on the global average tie-line power and the dual variable to obtain the expected tie-line switching power of this micronetwork includes: An augmented Lagrangian function is constructed based on the objective function, the global average tie-line power, and the dual variables. The augmented Lagrange function is: ; In the formula, Indicates the first The augmented Lagrangian function of a microgrid; This represents the local overall operating cost of this micronetwork; Let represent the vector of local decision variables for the m-th microgrid; Indicates the ADMM penalty parameter; This represents the tie-line switching power variable to be optimized. It represents the residual power determined by the global average tie-line power and the tie-line exchange power of the m-th microgrid in the k-th iteration; Let represent the dual variable of the k-th iteration; k represents the iteration number. With the goal of minimizing the augmented Lagrange function, the updated expected switching power of the tie lines in this micronetwork is obtained by solving the problem.

8. A distributed shared scheduling method for multiple micronets, characterized in that, Applied to the coordination center side, including: Send the initial global average tie-line power and dual variables to each microgrid; The expected switching power of the tie lines is obtained by solving the local optimization scheduling model of each microgrid based on the global average tie line power and the dual variable. The global average tie-line power is updated based on the received tie-line expected switching power of each micronet. The dual variable is updated based on the updated global average tie-line power. The original residual is determined based on the expected switching power of the contact lines of each microgrid, and it is determined whether the original residual is less than a preset accuracy threshold. If the original residual is less than the preset accuracy threshold, a stop iteration command is sent to each microgrid. If the original residual is greater than or equal to the preset accuracy threshold, then based on the updated global average tie-line power and the updated dual variable signal as the current round signal, the global average tie-line power and dual variable are sent to each microgrid based on the current round signal, and the process of solving the local optimization scheduling model of each microgrid to obtain the expected tie-line exchange power is returned.

9. The multi-micronet distributed shared scheduling method according to claim 8, characterized in that, The step of updating the global average tie-line power based on the received tie-line expected switching power of each micronet includes: The updated global average tie-line power is obtained by taking the arithmetic mean of the expected switching power of each microgrid tie-line. Correspondingly, updating the dual variable based on the updated global average tie-line power includes: Add the current dual variable to the updated global average tie-line power to obtain the updated dual variable; Correspondingly, determining the original residual based on the received expected switching power of the tie lines of each micronet includes: The norm of the sum of the expected switching power of each microgrid's tie lines is obtained based on the expected switching power of each microgrid's tie lines received. The norm is used as the original residual.

10. A multi-micronet distributed shared scheduling device, characterized in that, Applied to various microgrid sides, including: The acquisition module is used to acquire the operation data of this micronetwork and the charging behavior data of each electric vehicle within this micronetwork; The target charging curve construction module is used to construct a target state of charge curve based on the charging behavior data and using a preset range anxiety mapping relationship. The range anxiety quantification module is used to generate a deviation variable characterizing the user's range anxiety based on the target state of charge curve and the actual state of charge of the electric vehicle battery, and to obtain the range anxiety penalty cost based on the deviation variable. The energy storage degradation cost quantification module is used to determine the dynamic power boundary and degradation cost of the fixed energy storage system within this microgrid based on the operating data. The local cost quantification module is used to construct a local optimized scheduling model that includes the dynamic power boundary with the goal of minimizing the local comprehensive operating cost; wherein, the local comprehensive operating cost includes: net electricity purchase cost, range anxiety penalty cost, and degradation cost; The receiving module is used to receive the global average tie-line power and dual variable signals sent by the coordination center; The optimization module is used to solve the local optimization scheduling model based on the global average tie-line power and the dual variable to obtain the expected tie-line switching power of this micronetwork. The update module is used to send the expected switching power of the tie line to the coordination center, so that the coordination center updates the global average tie line power and the dual variable according to the expected switching power of the tie line, and returns to the step of receiving the global average tie line power and dual variable signal sent by the coordination center, until the convergence condition is met. The output module is used to receive the stop iteration command issued by the coordination center and output the local decision variables of this micronetwork as scheduling instructions.

Images