An intelligent peak shaving and valley filling calculation system and calculation method based on AOE network

Through the intelligent peak-cutting and valley-filling computing system based on the AOE network, combined with the linear planning model, the calculation of charge and discharge power is optimized, and the problems of traditional energy storage control strategies are not intelligent enough and cost-effective, realizing automatic adjustment and efficient peak-to-valley price difference arbitrage .

CN114400693BActive Publication Date: 2024-11-05WEISHI INTERACTIVE TECH CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202210052583.7
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2022-01-18
Publication Date
2024-11-05
Estimated Expiration
2042-01-18

AI Technical Summary

Technical Problem

The traditional user-side energy storage charging and discharging control strategy is not intelligent enough to adapt to changes in external conditions, and the implementation cost is high, which requires long-term coordination and adjustment by electrical and programmers.

Method used

An intelligent peak-cutting and valley-filling computing system based on AOE network is adopted, through actions and instructions between nodes, combined with a linear planning model, the charging and discharging power is optimized to achieve automatic adjustment and low-code control.

Benefits of technology

An intelligent adaptive charging and discharging strategy is realized, and the optimal charging and discharging power is calculated based on the SOC and electricity price situation every half hour, reducing the implementation cost and human resource requirements, and improving the arbitrage efficiency of peak and valley price difference.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN114400693B_ABST
    Figure CN114400693B_ABST
Patent Text Reader

Abstract

The present invention provides an intelligent peak shaving and valley filling calculation system and calculation method based on AOE (Activity on edge) network. The present invention constructs an intelligent peak shaving and valley filling calculation system based on AOE (Activity on edge) network. Based on this system, an intelligent peak shaving and valley filling strategy is implemented. This strategy calculates the optimal charging and discharging power at intervals based on future electricity price data, current energy storage SOC (State of Charge), maximum charging and discharging power, operating SOC range and other data, and obtains the optimal peak shaving and valley filling benefit.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention belongs to the technical field of smart grids, and in particular relates to an intelligent peak shaving and valley filling calculation system and a calculation method based on an AOE network. Background Art

[0002] User-side energy storage can realize arbitrage of peak-valley price difference by charging during the valley period of electricity price and discharging during the peak period of electricity price. The traditional charging and discharging control strategy is to analyze the electricity price data, manually determine the charging and discharging time, and then implement the charging and discharging strategy through code. This method has two disadvantages. First, it is not smart enough. When external conditions change, the strategy cannot be adjusted intelligently and adaptively. For example, when the electricity price changes in summer, or the load power changes, the discharge power must also change. When these situations occur, the charging and discharging strategy needs to be manually readjusted. Second, the implementation cost is high. Implementing a strategy requires the cooperation of electrical personnel and professional programmers, and it takes a long period of time. Summary of the invention

[0003] The present invention aims to solve the above technical problems and provides an intelligent peak shaving and valley filling calculation system and calculation method based on AOE network.

[0004] In order to achieve the above object, the present invention adopts the following technical solution:

[0005] An intelligent peak-shaving and valley-filling computing system based on an AOE network includes nodes A1-A9:

[0006] Node A1: Determine whether the SOC is in a reasonable range and the time is half an hour or the strategy has just started;

[0007] Action from node A1 to node A2: Divide 24 hours into 48 points of 0.5 hours, and calculate the current time as the time point of today;

[0008] Node A2: Determine whether the PCS is in the power-on state;

[0009] Action from node A2 to node A4: If the PCS is in the power-on state, this action is executed, otherwise there is no action;

[0010] Action from node A2 to node A3: If the PCS is not powered on, this action is executed and a PCS power-on command is sent;

[0011] Node A3: Determine whether the first PCS startup is successful;

[0012] Action from node A3 to node A4: If the first PCS startup is successful, this action is executed, otherwise there is no action;

[0013] Actions from node A3 to node A10: If the first PCS startup fails, perform a second PCS startup and send a PCS startup command;

[0014] Node A10: true regardless of whether the second PCS startup is successful;

[0015] Action from node A10 to node A4: No action;

[0016] Node A4: Start optimization calculation, and the judgment condition is always true;

[0017] Action from node A4 to node A5: Perform optimization calculation every half hour to calculate the current optimal charge and discharge power based on the current energy storage SOC, rated charge and discharge power, and electricity price in the next 48 hours;

[0018] Node A5: Determine whether the optimization calculation is successful;

[0019] Action from node A5 to node A6: If the optimization calculation is successful, the charging and discharging power is obtained;

[0020] Node A6: always true;

[0021] Actions from node A6 to node A8: Send charging and discharging power instructions to PCS;

[0022] Node A8: always true;

[0023] Actions from node A5 to node A7: If the optimization calculation fails, the default charge and discharge instructions are issued;

[0024] Node A7: always true;

[0025] Actions from node A7 to node A9: Update the value of the time variable;

[0026] Action from node A8 to node A9: Update the value of the time variable;

[0027] Node A9: Determine whether the time variable is updated successfully.

[0028] As a preferred technical solution, the intelligent peak shaving and valley filling computing system includes a SOC cross-border protection strategy model, and the SOC cross-border protection strategy model includes nodes B1-B8:

[0029] Node B1: Determine whether the SOC exceeds a reasonable range or reaches a boundary value;

[0030] Action from Node B1 to Node B2: No action;

[0031] Action from Node B1 to Node B3: No action;

[0032] Node B2: Determine whether the SOC exceeds a reasonable range;

[0033] Action from Node B2 to Node B4: No action;

[0034] Node B4: determines whether the PCS is in the power-on state;

[0035] Action from Node B4 to Node B6: If the PCS is in the power-on state, this action is performed, otherwise there is no action;

[0036] Action from Node B4 to Node B5: If the PCS is in a non-powered state, this action is executed and a PCS power-on instruction is sent;

[0037] Node B5: Determine whether the first PCS startup is successful;

[0038] Action from Node B5 to Node B6: If the first PCS startup is successful, this action is performed, otherwise there is no action;

[0039] Action from Node B5 to Node B8: If the first PCS startup fails, this action is executed and a second PCS startup instruction is sent;

[0040] Node B8: true regardless of whether the second PCS startup is successful;

[0041] Action from Node B8 to Node B6: No action;

[0042] Node B6: boot completed, always true;

[0043] Actions from node B6 to node B7: Send a rated reverse power command. If the SOC is less than the reasonable range, charge. If the SOC is greater than the reasonable range, discharge. Update the time variable.

[0044] Node B3: Determine whether the SOC reaches the boundary value or is abnormal;

[0045] Actions of node B3 to node B7: Send a stop charge and discharge instruction and update the time variable;

[0046] Node B7: Determine whether the time variable has been updated.

[0047] As a preferred technical solution, the intelligent peak shaving and valley filling calculation system also includes one or more of the following strategy models: over-temperature protection strategy model, over-voltage protection strategy model, fire protection strategy model, load protection strategy model;

[0048] The over-temperature protection strategy model includes nodes C1 and C2:

[0049] Node C1: Determine whether the maximum temperature of the single cell is greater than the set maximum threshold;

[0050] Action from node C1 to node C2: If the temperature is too high, a command to stop charging and discharging is issued;

[0051] Node C2: Determine whether charging and discharging has stopped;

[0052] The over-temperature protection strategy model includes nodes D1 and D2:

[0053] Node D1: Determine whether the maximum voltage of the single battery is greater than the set maximum threshold, and whether the minimum voltage is less than the set minimum threshold;

[0054] Action from node D1 to node D2: If overvoltage occurs, a command to stop charging and discharging is issued;

[0055] Node D2: Determine whether charging and discharging has stopped;

[0056] The fire protection strategy model includes nodes E1 and E2:

[0057] Node E1: Determine whether the fire alarm device is triggered;

[0058] Action from node E1 to node E2: If the fire alarm device is triggered, a command to stop charging and discharging is issued;

[0059] Node E2: Determine whether charging and discharging has stopped;

[0060] The load protection strategy model includes nodes F1 and F2:

[0061] Node F1: Determine whether the energy storage discharge power is greater than 80% of the system load;

[0062] Action from node F1 to node F2: If the discharge power is greater than 80% of the load, a stop charge and discharge instruction is issued;

[0063] Node F2: Determine whether charging and discharging has stopped.

[0064] AOE network: Activity on edge network. If in a weighted directed graph, vertices represent events, directed edges represent activities, and the weights on the edges represent the cost of the activity (such as the duration of the activity), then this weighted directed graph is called an AOE network. If an AOE network is used to represent a project, it is not enough to just consider the priority relationship between the various sub-projects. More attention should be paid to the shortest time to complete the entire project; which activities will affect the progress of the entire project if they are postponed, and whether accelerating these activities will improve the efficiency of the entire project. Therefore, the activities required to complete the scheduled project plan, the time when each activity is planned to be completed, which events are to occur, and the relationship between these events and activities are usually listed in the AOE network, so as to determine whether the project is feasible, estimate the time to complete the project, and determine which activities are the key to affecting the progress of the project.

[0065] AOE network has the following two properties:

[0066] ① Only after the event represented by a vertex occurs, the activities represented by the directed edges starting from the vertex can begin.

[0067] ②The event represented by a vertex can only occur when all the activities represented by the directed edges entering a vertex have ended.

[0068] An intelligent peak shaving and valley filling calculation method based on AOE network includes the following steps:

[0069] Step 1: determine whether the energy storage SOC is in a reasonable range, and the time is half an hour or the strategy has just started. If so, execute step 2;

[0070] Step 2: Divide 24 hours into 48 points of 0.5 hours, and calculate the current time as the time point of today;

[0071] Step 3, determine whether the energy storage converter PCS is in the power-on state, if the PCS is in the power-on state, no action is performed, if the PCS is in the non-power-on state, send a PCS power-on instruction;

[0072] Step 4, determine whether the first PCS startup is successful. If the first PCS startup is successful, no action is performed. If the first PCS startup is unsuccessful, perform a second PCS startup and send a PCS startup command.

[0073] Step 5: Perform optimization calculation every half an hour to calculate the current optimal charge and discharge power based on the current energy storage SOC, rated charge and discharge power, and electricity prices in the next 48 hours;

[0074] Step 6, determine whether the optimization calculation is successful. If the optimization calculation is successful, obtain the charge and discharge power. If the optimization calculation fails, issue a default charge and discharge instruction.

[0075] Step 7, update the value of the time variable;

[0076] Step 8: Determine whether the time variable is updated successfully.

[0077] As a preferred technical solution, in step 5, the method for solving the optimal charging and discharging power includes:

[0078] Construct a linear programming model for the charging and discharging power of the energy storage system, including:

[0079] (1) Define the objective function:

[0080]

[0081] Convert to find the minimum value:

[0082]

[0083] Among them, suppose the energy storage discharge power ui at time i>=0, suppose the energy storage charging power vi at time i>=0, and the electricity price at time i (yuan / kwh): PriceElce i , Single charge and discharge cycle cost (yuan / time): PriceCycle, Initial capacity (kwh): Q0 = QRate * SOC0, Rated power (kw): Prate, Rated capacity (kwh): QRate, Minimum discharge SOC (%): DOD1, Maximum charge SOC (%): DOD2, Data collection interval (hours): Δt, Charge and discharge conversion efficiency: CDRate, Battery remaining capacity: SOH;

[0084] (2) Define constraints:

[0085] a. The capacity of energy storage at each moment is greater than the rated capacity multiplied by the remaining discharge SOC:

[0086] (u1-v1*CDRate)*Δt<=Q0-DOD1*QRate

[0087] (u1-v1*CDRate)*Δt+(u2-v2*CDRate)*Δt<=Q0-DOD1*QRate ....

[0089] (u1-v1*CDRate)*Δt+...+(u n -v n *CDRate)*Δt<=Q0-DOD1*QRate

[0090] b. The capacity of energy storage at each moment is less than the rated capacity:

[0091] -(u1-v1*CDRate)*Δt<=QRate*DOD2-Q0

[0092] -(u1-v1*CDRate)*Δt-(u2-v2*CDRate)*Δt<=QRate*DOD2-Q0 ....

[0094] -(u1-v1*CDRate)*Δt-...-(u n -v n *CDRate)*Δt<=QRate*DOD2-Q0;

[0095] c. Limitation of the size range of decision variables

[0096] 0<=v i <=PRate

[0097] 0<=u i <=PRate.

[0098] As a preferred technical solution, the intelligent peak shaving and valley filling calculation method based on the AOE network also includes one or more of the following protection strategies: over-temperature protection strategy, over-voltage protection strategy, fire protection strategy, and load protection strategy;

[0099] The SOC out-of-bounds protection strategy includes: in step 1, if the SOC exceeds the reasonable range, then execute step 3 and step 4 in sequence. After the PCS is turned on, issue a rated reverse power instruction. If the SOC is less than the reasonable range, charge. If the SOC is greater than the reasonable range, discharge. If the SOC reaches the boundary value or is abnormal, issue a stop charging instruction and update the time variable. Finally, execute step 7 and step 8 in sequence.

[0100] The over-temperature protection strategy includes: determining whether the maximum temperature of the single battery is greater than the set maximum threshold, and if it is over-temperature, issuing a stop charge and discharge instruction; then determining whether the charge and discharge have been stopped;

[0101] The overvoltage protection strategy includes: judging whether the maximum voltage of the single battery is greater than the set maximum threshold, and whether the minimum voltage is less than the set minimum threshold. If overvoltage occurs, a stop charge and discharge instruction is issued; and then judging whether charge and discharge have been stopped.

[0102] The fire protection strategy includes: determining whether the fire alarm device is triggered, and if the fire alarm is triggered, issuing a stop charge and discharge instruction; and then determining whether the charge and discharge have been stopped;

[0103] The load protection strategy includes: judging whether the energy storage discharge power is greater than 80% of the system load, and if the discharge power is greater than 80% of the load, issuing a stop charge and discharge instruction; and then judging whether the charge and discharge have been stopped.

[0104] After adopting the above technical solution, the present invention has the following advantages:

[0105] The present invention constructs an intelligent peak shaving and valley filling calculation system based on the AOE (Activity on edge) network. Based on this system, an intelligent peak shaving and valley filling strategy is implemented. This strategy calculates the optimal charging and discharging power at regular intervals based on future electricity price data, current energy storage SOC (State of Charge), maximum charging and discharging power, operating SOC range and other data, and obtains the optimal peak shaving and valley filling benefit. BRIEF DESCRIPTION OF THE DRAWINGS

[0106] Figure 1 It is a structural diagram of an intelligent peak-shaving and valley-filling computing system based on AOE network;

[0107] Figure 2 It is a structural diagram of the SOC cross-border protection strategy model;

[0108] Figure 3 It is a structural schematic diagram of an over-temperature protection strategy model, an over-voltage protection strategy model, a fire protection strategy model, and a load protection strategy model;

[0109] Figure 4 The energy storage SOC and power curve after applying the intelligent peak-shaving and valley-filling AOE strategy for a certain period of time on a certain day. DETAILED DESCRIPTION

[0110] The present invention is further described in detail below in conjunction with the accompanying drawings and specific embodiments.

[0111] like Figure 1 As shown, an intelligent peak shaving and valley filling computing system based on AOE network includes nodes A1-A9:

[0112] Node A1: Determine whether the SOC is in a reasonable range and the time is half an hour or the strategy has just started;

[0113] Action from node A1 to node A2: Divide 24 hours into 48 points of 0.5 hours, and calculate the current time as the time point of today;

[0114] Node A2: Determine whether the PCS is in the power-on state;

[0115] Action from node A2 to node A4: If the PCS is in the power-on state, this action is executed, otherwise there is no action;

[0116] Action from node A2 to node A3: If the PCS is not powered on, this action is executed and a PCS power-on command is sent;

[0117] Node A3: Determine whether the first PCS startup is successful;

[0118] Action from node A3 to node A4: If the first PCS startup is successful, this action is executed, otherwise there is no action;

[0119] Actions from node A3 to node A10: If the first PCS startup fails, perform a second PCS startup and send a PCS startup command;

[0120] Node A10: true regardless of whether the second PCS startup is successful;

[0121] Action from node A10 to node A4: No action;

[0122] Node A4: Start optimization calculation, and the judgment condition is always true;

[0123] Action from node A4 to node A5: Perform optimization calculation every half hour to calculate the current optimal charge and discharge power based on the current energy storage SOC, rated charge and discharge power, and electricity price in the next 48 hours;

[0124] Node A5: Determine whether the optimization calculation is successful;

[0125] Action from node A5 to node A6: If the optimization calculation is successful, the charging and discharging power is obtained;

[0126] Node A6: always true;

[0127] Actions from node A6 to node A8: Send charging and discharging power instructions to PCS;

[0128] Node A8: always true;

[0129] Actions from node A5 to node A7: If the optimization calculation fails, the default charge and discharge instructions are issued;

[0130] Node A7: always true;

[0131] Actions from node A7 to node A9: Update the value of the time variable;

[0132] Action from node A8 to node A9: Update the value of the time variable;

[0133] Node A9: Determine whether the time variable is updated successfully.

[0134] like Figure 2As shown, the intelligent peak shaving and valley filling computing system includes a SOC cross-boundary protection strategy model, and the SOC cross-boundary protection strategy model includes nodes B1-B8:

[0135] Node B1: Determine whether the SOC exceeds a reasonable range or reaches a boundary value;

[0136] Action from Node B1 to Node B2: No action;

[0137] Action from Node B1 to Node B3: No action;

[0138] Node B2: Determine whether SOC exceeds the reasonable range; if the reasonable range of SOC is 0.1<=SOC<=0.9, the unreasonable range is [0,0.1) and (0.9,1];

[0139] Action from Node B2 to Node B4: No action;

[0140] Node B4: determines whether the PCS is in the power-on state;

[0141] Action from Node B4 to Node B6: If the PCS is in the power-on state, this action is performed, otherwise there is no action;

[0142] Action from Node B4 to Node B5: If the PCS is in a non-powered state, this action is executed and a PCS power-on instruction is sent;

[0143] Node B5: Determine whether the first PCS startup is successful;

[0144] Action from Node B5 to Node B6: If the first PCS startup is successful, this action is performed, otherwise there is no action;

[0145] Action from Node B5 to Node B8: If the first PCS startup fails, this action is executed and a second PCS startup instruction is sent;

[0146] Node B8: true regardless of whether the second PCS startup is successful;

[0147] Action from Node B8 to Node B6: No action;

[0148] Node B6: boot completed, always true;

[0149] Actions from node B6 to node B7: Send the rated reverse power command. If the SOC is less than the reasonable range, charge. If the SOC is greater than the reasonable range, discharge, and update the time variable. If the SOC is in [0, 0.1), charge. If the SOC is in (0.9, 1], discharge, and update the time variable.

[0150] Node B3: Determine whether the SOC reaches the boundary value or is abnormal;

[0151] Actions of node B3 to node B7: Send a stop charge and discharge instruction and update the time variable;

[0152] Node B7: Determine whether the time variable has been updated.

[0153] like Figure 3 As shown, the intelligent peak shaving and valley filling calculation system also includes one or more of the following strategy models: over-temperature protection strategy model, over-voltage protection strategy model, fire protection strategy model, load protection strategy model;

[0154] The over-temperature protection strategy model includes nodes C1 and C2:

[0155] Node C1: Determine whether the maximum temperature of the single cell is greater than the set maximum threshold;

[0156] Action from node C1 to node C2: If the temperature is too high, a command to stop charging and discharging is issued;

[0157] Node C2: Determine whether charging and discharging has stopped;

[0158] The over-temperature protection strategy model includes nodes D1 and D2:

[0159] Node D1: Determine whether the maximum voltage of the single battery is greater than the set maximum threshold, and whether the minimum voltage is less than the set minimum threshold;

[0160] Action from node D1 to node D2: If overvoltage occurs, a command to stop charging and discharging is issued;

[0161] Node D2: Determine whether charging and discharging has stopped;

[0162] The fire protection strategy model includes nodes E1 and E2:

[0163] Node E1: Determine whether the fire alarm device is triggered;

[0164] Action from node E1 to node E2: If the fire alarm device is triggered, a command to stop charging and discharging is issued;

[0165] Node E2: Determine whether charging and discharging has stopped;

[0166] The load protection strategy model includes nodes F1 and F2:

[0167] Node F1: Determine whether the energy storage discharge power is greater than 80% of the system load;

[0168] Action from node F1 to node F2: If the discharge power is greater than 80% of the load, a stop charge and discharge instruction is issued;

[0169] Node F2: Determine whether charging and discharging has stopped.

[0170] AOE network: Activity on edge network. If in a weighted directed graph, vertices represent events, directed edges represent activities, and the weights on the edges represent the cost of the activity (such as the duration of the activity), then this weighted directed graph is called an AOE network. If an AOE network is used to represent a project, it is not enough to just consider the priority relationship between the various sub-projects. More attention should be paid to the shortest time to complete the entire project; which activities will affect the progress of the entire project if they are postponed, and whether accelerating these activities will improve the efficiency of the entire project. Therefore, the activities required to complete the scheduled project plan, the time when each activity is planned to be completed, which events are to occur, and the relationship between these events and activities are usually listed in the AOE network, so as to determine whether the project is feasible, estimate the time to complete the project, and determine which activities are the key to affecting the progress of the project.

[0171] AOE network has the following two properties:

[0172] ① Only after the event represented by a vertex occurs, the activities represented by the directed edges starting from the vertex can begin.

[0173] ②The event represented by a vertex can only occur when all the activities represented by the directed edges entering a vertex have ended.

[0174] This intelligent peak shaving and valley filling calculation system is implemented based on a low-code controller platform.

[0175] The low-code controller uses an event-driven approach to design complex control strategies, models the control strategies through the AOE network, and has the functions of existing programmable controllers such as logical operations, sequential control, timing, counting, and arithmetic operations. It also supports equation solving and optimization model solving functions, and can be applied to many control scenarios in the power industry, such as integrated energy system control, microgrid control, and energy storage control. In addition, for large systems, such as regional distribution networks, large-scale distributed real-time control can be achieved through the collaboration of multiple low-code controllers based on edge computing technology. Low-code controllers require almost no code writing. Users only need to define and describe the AOE network corresponding to the control strategy in a given file format to complete the design of the control link. Therefore, it is more convenient for people with non-programming technical backgrounds to use. On the other hand, the control strategy implementation method adopted by the low-code controller is universal, and it is easier to design by analogy than traditional controllers.

[0176] In the actual application scenarios of new power systems, low-code controllers are more reliable industrial control devices.

[0177] It has the technical advantages of more intuitive control strategy conversion, simpler control link design, more universal control model establishment, and more efficient control target realization:

[0178] (1) More reliable industrial control equipment

[0179] The low-code controller fully considers various interference factors and incidental events. When the system is abnormal, timely measures are taken to correct the non-permanent errors of the system, try to restore the system to normal, and ensure the correctness and continuity of control. This real-time diagnostic capability of the low-code controller ensures its high reliability in industrial control scenarios. Specifically, the low-code controller adds failure mode, timeout and other settings to the AOE network design. During the execution process, the execution time and execution reliability of the network are controlled according to the flexible timeout settings; the failure modes of the action include retry, ignore, stop the affected nodes, and stop the entire AOE network. Several levels can be flexibly set according to the importance of the action.

[0180] (2) More intuitive control strategy conversion

[0181] When the model is converted into a control strategy, the model is converted into an AOE network structure, which includes variables, events and actions. There is a logical connection between the design of events and the actions executed after the events are triggered. The control strategy can be represented graphically through the AOE network. Compared with conversion into code language, this conversion process is more vivid and intuitive, and less prone to errors during implementation.

[0182] (3) More standard control link design

[0183] When designing and implementing specific control links, low-code controllers require almost no code writing. Instead, they only need to write relevant configuration files, which are completed in a standardized format by entering information in tables, greatly shortening the control design process.

[0184] (4) Establishment of a more general control model

[0185] The low-code controller provides a universal framework, which has two aspects:

[0186] A. The design of the control strategy has universal characteristics. The use of this low-code controller is easier to familiarize and master than the traditional controller. You only need to complete the control target of one model, and then you can refer to this model to complete the control target of other models;

[0187] B. The control framework can run on ARM and x86 architecture processors, and supports multiple operating systems such as Linux, Windows, and MacOS. It supports running on low-performance hardware such as Raspberry Pi and has the ability to run across platforms.

[0188] (5) More efficient control target realization

[0189] On the one hand, event-driven control is used as the basis for AOE network execution. When an event occurs due to a change in the measurement point, the controller updates its own status and communicates. Compared with the time-driven control method, this mechanism can reduce the waste of communication resources and energy consumption caused by unnecessary communication, and run under hardware with low communication conditions. On the other hand, for a control target, it is only necessary to build a suitable AOE control network to realize the function of a control system without the need to set up complicated control links in partition blocks. The function is realized more clearly and faster.

[0190] Based on the above computing system, this embodiment also provides an intelligent peak shaving and valley filling computing method based on an AOE network, comprising the following steps:

[0191] Step 1: determine whether the energy storage SOC is in a reasonable range, and the time is half an hour or the strategy has just started. If so, execute step 2;

[0192] Step 2: Divide 24 hours into 48 points of 0.5 hours, and calculate the current time as the time point of today;

[0193] Step 3, determine whether the energy storage converter PCS is in the power-on state, if the PCS is in the power-on state, no action is performed, if the PCS is in the non-power-on state, send a PCS power-on instruction;

[0194] Step 4, determine whether the first PCS startup is successful. If the first PCS startup is successful, no action is performed. If the first PCS startup is unsuccessful, perform a second PCS startup and send a PCS startup command.

[0195] Step 5: Perform optimization calculation every half an hour to calculate the current optimal charge and discharge power based on the current energy storage SOC, rated charge and discharge power, and electricity prices in the next 48 hours;

[0196] Step 6, determine whether the optimization calculation is successful. If the optimization calculation is successful, obtain the charge and discharge power. If the optimization calculation fails, issue a default charge and discharge instruction.

[0197] Step 7, update the value of the time variable;

[0198] Step 8: Determine whether the time variable is updated successfully.

[0199] In step 5, the method for solving the optimal charging and discharging power includes:

[0200] Construct a linear programming model for the charging and discharging power of the energy storage system, including:

[0201] (1) Define the objective function:

[0202]

[0203] Convert to find the minimum value:

[0204]

[0205] Among them, suppose the energy storage discharge power ui at time i>=0, suppose the energy storage charging power vi at time i>=0, and the electricity price at time i (yuan / kwh): PriceElce i , Single charge and discharge cycle cost (yuan / time): PriceCycle, Initial capacity (kwh): Q0 = QRate * SOC0, Rated power (kw): Prate, Rated capacity (kwh): QRate, Minimum discharge SOC (%): DOD1, Maximum charge SOC (%): DOD2, Data collection interval (hours): Δt, Charge and discharge conversion efficiency: CDRate, Battery remaining capacity: SOH;

[0206] (2) Define constraints:

[0207] a. The capacity of energy storage at each moment is greater than the rated capacity multiplied by the remaining discharge SOC:

[0208] (u1-v1*CDRate)*Δt<=Q0-DOD1*QRate

[0209] (u1-v1*CDRate)*Δt+(u2-v2*CDRate)*Δt<=Q0-DOD1*QRate ....

[0211] (u1-v1*CDRate)*Δt+...+(u n -v n *CDRate)*Δt<=Q0-DOD1*QRate;

[0212] b. The capacity of energy storage at each moment is less than the rated capacity:

[0213] -(u1-v1*CDRate)*Δt<=QRate*DOD2-Q0

[0214] -(u1-v1*CDRate)*Δt-(u2-v2*CDRate)*Δt<=QRate*DOD2-Q0 ....

[0216] -(u1-v1*CDRate)*Δt-...-(u n -v n *CDRate)*Δt<=QRate*DOD2-Q0;

[0217] c. Limitation of the size range of decision variables

[0218] 0<=v i <=PRate

[0219] 0<=u i <=PRate.

[0220] The intelligent peak shaving and valley filling calculation method based on the AOE network also includes one or more of the following protection strategies: over-temperature protection strategy, over-voltage protection strategy, fire protection strategy, and load protection strategy;

[0221] The SOC out-of-bounds protection strategy includes: in step 1, if the SOC exceeds the reasonable range, then execute step 3 and step 4 in sequence. After the PCS is turned on, issue a rated reverse power instruction. If the SOC is less than the reasonable range, charge. If the SOC is greater than the reasonable range, discharge. If the SOC reaches the boundary value or is abnormal, issue a stop charging instruction and update the time variable. Finally, execute step 7 and step 8 in sequence.

[0222] The over-temperature protection strategy includes: determining whether the maximum temperature of the single battery is greater than the set maximum threshold, and if it is over-temperature, issuing a stop charge and discharge instruction; then determining whether the charge and discharge have been stopped;

[0223] The overvoltage protection strategy includes: judging whether the maximum voltage of the single battery is greater than the set maximum threshold, and whether the minimum voltage is less than the set minimum threshold. If overvoltage occurs, a stop charge and discharge instruction is issued; and then judging whether charge and discharge have been stopped.

[0224] The fire protection strategy includes: determining whether the fire alarm device is triggered, and if the fire alarm is triggered, issuing a stop charge and discharge instruction; and then determining whether the charge and discharge have been stopped;

[0225] The load protection strategy includes: judging whether the energy storage discharge power is greater than 80% of the system load, and if the discharge power is greater than 80% of the load, issuing a stop charge and discharge instruction; and then judging whether the charge and discharge have been stopped.

[0226] In this embodiment, this intelligent peak-shaving and valley-filling calculation system and calculation method are applied in a data center energy storage project. The energy storage capacity is 1100Kwh, the maximum charging power is 200Kw, and the maximum discharging power is 240Kw. The non-summer time-of-use electricity price is as follows:

[0227] time 22-6 6-8 8-11 11-18 18-21 21-22 Electricity Prices 0.256 0.494 0.843 0.494 0.843 0.494

[0228] Figure 4 The energy storage SOC and power curve after applying the intelligent peak-shaving and valley-filling AOE strategy for a certain period of time on a certain day.

[0229] In addition to the above preferred embodiments, the present invention has other implementation modes. Those skilled in the art may make various changes and modifications based on the present invention. As long as they do not depart from the spirit of the present invention, they should all fall within the scope defined by the claims attached to the present invention.

Claims

1. An intelligent peak shaving and valley filling calculation method based on AOE network, characterized in that: The following steps are involved: Step 1: determine whether the energy storage SOC is in a reasonable range, and the time is half an hour or the strategy has just started. If so, execute step 2; Step 2: Divide 24 hours into 48 points of 0.5 hours, and calculate the current time as the time point of today; Step 3, determine whether the energy storage converter PCS is in the power-on state, if the PCS is in the power-on state, no action is performed, if the PCS is in the non-power-on state, send a PCS power-on instruction; Step 4, determine whether the first PCS startup is successful. If the first PCS startup is successful, no action is performed. If the first PCS startup is unsuccessful, perform a second PCS startup and send a PCS startup command. Step 5: Perform optimization calculation every half an hour to calculate the current optimal charge and discharge power based on the current energy storage SOC, rated charge and discharge power, and electricity prices in the next 48 hours; Step 6, determine whether the optimization calculation is successful. If the optimization calculation is successful, obtain the charge and discharge power. If the optimization calculation fails, issue a default charge and discharge instruction. Step 7, update the value of the time variable; Step 8, determine whether the time variable is updated successfully; In step 5, the method for solving the optimal charging and discharging power includes: Construct a linear programming model for the charging and discharging power of the energy storage system, including: (1) Define the objective function: Convert to find the minimum value: Among them, suppose the energy storage discharge power ui>=0 at time i, suppose the energy storage charging power vi>=0 at time i, and the electricity price at time i, in units of yuan / kwh: PriceElce i , Single charge and discharge cycle cost, unit is yuan / time: PriceCycle, Initial capacity, unit is kwh: Q0=QRate*SOC0, Rated power, unit is kw: Prate, Rated capacity, unit is kwh: QRate, Minimum discharge SOC, in percentage: DOD1, Maximum charge SOC, in percentage: DOD2, Data collection interval, in hours: Δt, Charge and discharge conversion efficiency: CDRate, Remaining battery capacity: SOH; (2) Define constraints: a. The capacity of energy storage at each moment is greater than the rated capacity multiplied by the remaining discharge SOC: (u1-v1*CDRate)*Δt<=Q0-DOD1*QRate (u1-v1*CDRate)*Δt+(u2-v2*CDRate)*Δt<=Q0-DOD1*QRate .... (u1-v1*CDRate)*Δt+...+(u n -v n *CDRate)*Δt<=Q0-DOD1*QRate b. The capacity of energy storage at each moment is less than the rated capacity: -(u1-v1*CDRate)*Δt<=QRate*DOD2-Q0 -(u1-v1*CDRate)*Δt-(u2-v2*CDRate)*Δt<=QRate*DOD2-Q0 .... -(u1-v1*CDRate)*Δt-...-(u n -v n *CDRate)*Δt<=QRate*DOD2-Q0; c. Limitation of the size range of decision variables 0<=v i <=PRate 0<=u i <= PRate.

2. The intelligent peak shaving and valley filling calculation method based on AOE network as claimed in claim 1, characterized in that: The intelligent peak shaving and valley filling calculation method based on the AOE network also includes one or more of the following protection strategies: over-temperature protection strategy, over-voltage protection strategy, fire protection strategy, and load protection strategy; The SOC out-of-bounds protection strategy includes: in step 1, if the SOC exceeds the reasonable range, then execute step 3 and step 4 in sequence. After the PCS is turned on, issue a rated reverse power instruction. If the SOC is less than the reasonable range, charge. If the SOC is greater than the reasonable range, discharge. If the SOC reaches the boundary value or is abnormal, issue a stop charging instruction and update the time variable. Finally, execute step 7 and step 8 in sequence. The over-temperature protection strategy includes: determining whether the maximum temperature of the single battery is greater than the set maximum threshold, and if it is over-temperature, issuing a stop charge and discharge instruction; then determining whether the charge and discharge have been stopped; The overvoltage protection strategy includes: judging whether the maximum voltage of the single battery is greater than the set maximum threshold, and whether the minimum voltage is less than the set minimum threshold. If overvoltage occurs, a stop charge and discharge instruction is issued; then judging whether charge and discharge have been stopped; The fire protection strategy includes: determining whether the fire alarm device is triggered, and if the fire alarm is triggered, issuing a stop charge and discharge instruction; and then determining whether the charge and discharge have been stopped; The load protection strategy includes: judging whether the energy storage discharge power is greater than 80% of the system load, and if the discharge power is greater than 80% of the load, issuing a stop charge and discharge instruction; and then judging whether the charge and discharge have been stopped.