Distributed control method for large-scale energy storage power station

Abstract

The invention provides a distributed control method for a large-scale energy storage power station. The distributed control method comprises the following steps: S1, constructing a control architecture consisting of a unit control layer and an information acquisition layer; s2, designing a distributed frequency modulation control strategy based on a first-order discrete weighted consistency algorithm, and optimizing the algorithm by adopting a block matrix technology; s3, an Ethernet control automation master-slave communication system is established, the master station is responsible for power data calculation and output, and the master station and the slave station carry out real-time communication through an Ethernet protocol; and S4, designing a control program, setting the maximum communication waiting time, establishing client array management communication, and adjusting a data exchange period and a communication strategy according to requirements. Through a hierarchical distributed control architecture and a consistency algorithm, efficient energy storage unit control and power grid dispatching response are realized, and the performance bottleneck of centralized control is solved. Meanwhile, the provided Ethernet control automation technology optimizes the communication structure, meets the real-time and high-capacity requirements, and reduces the system cost.

Description

Technical Field

[0001] The present invention relates to the technical field of energy storage power stations, to the field of distributed control, and in particular to a distributed control method for a large-scale energy storage power station. Background Art

[0002] Benefiting from the booming development of electric vehicles, battery technology, especially lithium-ion batteries, has made breakthrough progress, battery energy efficiency has been significantly improved, and costs have also decreased. Influenced by this, the installed capacity of lithium-ion battery energy storage power stations has also grown in leaps and bounds, and has now gradually increased from megawatts to hundreds of megawatts and above. Since large-scale energy storage power stations are generally composed of multiple small-capacity battery units, the amount of data brought is huge, which puts high demands on the performance of the central processing unit of centralized control. Therefore, how to enable energy storage power stations to effectively and accurately control each unit while responding to the dispatching needs of the power grid with high precision is a problem that needs to be solved urgently.

[0003] In addition, the massive data collection and transmission processing needs of energy storage power stations require that the communication lines have a certain carrying capacity while ensuring data transmission accuracy, and will not collapse due to the large amount of data. The characteristics of energy storage power stations such as fast response speed and the needs of rapid adjustment of power systems also put forward higher requirements for the real-time and rapidity of energy storage power station communications. Currently, the field bus specifications used in industrial production include Modbus, Profibus, ControlNet and other protocols, but these protocols are difficult to meet the requirements of real-time and high capacity while pursuing low costs.

[0004] Therefore, it is necessary to invent a distributed control method for large-scale energy storage power stations to support the optimized operation control of large-scale energy storage power stations and provide solutions for power allocation strategies when large-scale energy storage power stations participate in frequency modulation, peak shaving and valley filling and other scenarios. Summary of the invention

[0005] To achieve the above object, the present invention provides a large-scale energy storage power station distributed control method, comprising the following steps:

[0006] Step S1: Building a distributed control architecture for a large-scale energy storage power station, the architecture includes a unit control layer and an information collection layer, wherein the unit control layer is responsible for local control and coordination of energy storage units in the energy storage power station, including at least one leading energy storage terminal and multiple subordinate energy storage terminals, and the information collection layer is responsible for real-time data collection and transmission;

[0007] Step S2: Design a distributed frequency regulation control strategy for the energy storage power station. The strategy is based on a first-order discrete weighted consensus algorithm and uses a block matrix technology to convert the constant term in the algorithm into a block matrix. By dividing each matrix into blocks, the consistency algorithm can adapt to the design of a first-order discrete weighted consensus algorithm in a hierarchical framework.

[0008] Step S3: Build an Ethernet control automation technology master-slave communication system for the energy storage power station, where the Ethernet control automation technology master station is the dominant energy storage terminal, responsible for calculating and outputting the power data information of the energy storage unit; the Ethernet control automation technology slave station is the subordinate energy storage terminal, responsible for receiving the data frames sent by the master station and executing the corresponding control instructions. The master station and the slave station achieve high-speed real-time communication through the Ethernet control automation technology protocol.

[0009] Step S4: Design and implement the Ethernet control automation technology program of the energy storage power station, including: setting the maximum waiting time for communication and monitoring the communication delay, establishing a client array to manage the communication of multiple energy storage units, formulating the data exchange cycle between the Ethernet control automation technology master station and the slave station, and dynamically adjusting the communication frequency and data processing strategy according to actual needs.

[0010] Furthermore, in step S1, the unit control layer includes multiple energy storage terminals, at least one of which is a dominant energy storage terminal, responsible for receiving superior scheduling instructions, starting the corresponding distributed algorithm, and sending iteration signals to other energy storage terminals; the unit control layer also includes multiple slave energy storage terminals connected to the energy storage units, responsible for data collection and control of the energy storage units, and information exchange based on the received iteration signals.

[0011] Furthermore, in step S1, if the slave energy storage terminal receives an end signal sent by the master energy storage terminal, the information interaction is stopped, and a control signal is generated according to the final iteration result to adjust the output power of the controlled energy storage unit.

[0012] Furthermore, in step S1, the information acquisition layer is responsible for connecting multiple battery packs, each battery pack is composed of multiple energy storage units through corresponding energy storage terminals, and the information acquisition layer collects and communicates data of each battery pack through each energy storage terminal.

[0013] Further, step S2 includes the following steps:

[0014] Step S201: construct a distributed frequency regulation control strategy model for a large-scale energy storage power station, which is expressed as shown in formula (1):

[0015]

[0016] Where: x is the state variable; k is the discrete time; H is the iteration matrix of the state variable; I n is the n-order unit matrix, n is the total number of energy storage terminals; ε is the set iteration step; L is the Laplace matrix; W is the weighted diagonal matrix based on the battery state of charge SOC value, and the specific expression is shown in formula (2):

[0017] W=diag(w1,..., w i ,...,w n ) (2)

[0018] Where: w i The calculation formulas are shown in equations (3) and (4), where equation (4) is used to obtain the relative weight of the battery state of charge SOC value:

[0019]

[0020] Where: SOC i is the battery state of charge SOC value currently collected by energy storage unit i; SOC max , SOC min They are the minimum and maximum values ​​of the state of charge (SOC) of the energy storage unit battery respectively;

[0021] Step S202: With the continuous iteration of the state variable x in step S201, the state variable x will gradually approach the weighted average value of the initial value x0, and the initial value x0 of the state variable x is calculated, and the expression is shown in formula (5):

[0022]

[0023] Where: P0 is an n×1 column matrix; P com is the active power dispatch value sent by the superior to the dominant node 1;

[0024] Step S203: Taking power as a state variable, combined with formula (1), the output power of the energy storage power station is subjected to distributed calculation, and the expression is shown in formula (6):

[0025]

[0026] Step S204: Based on the block matrix method, the constant term contained in equation (6) is converted into a block matrix:

[0027] The SOC of formula (3) i Replace it with an n×1 column matrix and replace SOC max and SOC min All are set to n×1 column matrices;

[0028] Replace the denominator of the division in formula (4) with And divide W into blocks so that each w i is a diagonal matrix, and its specific expression is shown in formula (7):

[0029]

[0030] In formula (1), I nL is divided into blocks, and each element in the matrix is ​​converted into a block matrix, with I n For example, the expression after block division is shown in formula (8):

[0031]

[0032] Replace P0 in formula (5) with n 2 ×1 column matrix, where P com Synchronously convert to n×1 column matrix;

[0033] Step S205: Considering the distribution problem between power and each energy storage unit, the weight coefficient of each energy storage unit is adjusted. The specific calculation formula is shown in formula (9):

[0034]

[0035] Furthermore, in step S3, the Ethernet control automation technology master station needs to perform the following operations:

[0036] Initialization operation: The master station completes the initialization operation of each slave station according to the upper layer instructions;

[0037] Instruction analysis and data frame generation: The master station analyzes the instruction content according to the received instruction, writes the corresponding data frame, and sends it to each controlled slave station;

[0038] Data analysis and visualization: The master station analyzes and processes the data returned by the slave station and displays the results in a visual way for developers’ reference;

[0039] Data upload: The master station uploads the data required by the superior according to the demand.

[0040] Further, in step S3, the Ethernet control automation technology slave device needs to perform the following operations:

[0041] Data communication: bidirectional communication with the master station, outputting the data output by the master station to the application, and receiving the input data from the application and sending it back to the master station;

[0042] Data frame processing and analysis: Process and analyze Ethernet control automation technology data frames received from the master station. When the data frame passes through the slave station, the slave station controller extracts and parses the control command, stores the data to be output into the storage area, and writes the input data into the corresponding sub-message;

[0043] Control device operation and data sampling: The slave control microprocessor reads the data in the slave controller storage area and outputs it to the corresponding device; at the same time, the slave control microprocessor samples the feedback data from the device and sends the feedback data back to the slave controller storage area as input data for storage.

[0044] Further, step S4 includes the following steps:

[0045] Step S401: Setting a maximum waiting time to prevent the program from being stuck due to communication failure. If the actual waiting time exceeds the set value, the loop is forced to exit and an error is reported to the user.

[0046] Step S402: In order to manage and coordinate the input and output of the slave station, the slave station sets a client array whose size is equal to the total number of slave stations, and each array element contains relevant information of a slave station. By calling the array, the input and output format of the slave station can be determined, so as to perform operations such as amplitude;

[0047] Step S403: After the master station energy storage terminal obtains the power allocation result through distributed computing, it formulates the operation cycle plan for data exchange between the master station and the slave station based on the polling cycle of the master station. If the data is sent in a cycle, it will cause unnecessary communication pressure. Therefore, the event-driven triggering method is used to trigger the data sending. The energy storage information transmitted by the slave station energy storage terminal is polled to ensure the real-time acquisition of the operation status of each energy storage unit.

[0048] The beneficial effects of the present invention are:

[0049] The present invention designs a hierarchical distributed large-scale energy storage power station architecture, and on this basis proposes a first-order discrete weighted consistency algorithm that considers the block matrix, and designs a distributed frequency modulation control strategy for the energy storage power station with the consistency algorithm as the core. This invention effectively solves the problem of centralized control's strong dependence on the performance of the central processor, allowing the energy storage power station to effectively and accurately control each unit while responding to the grid dispatching needs with high precision.

[0050] In addition, the present invention proposes a master-slave station construction and programming method for Ethernet control automation technology of energy storage power stations to realize the communication of energy storage power stations, so as to support the communication needs of distributed control of energy storage power stations from a practical engineering perspective, and meet the requirements of real-time and high capacity while pursuing low cost. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] Figure 1 It is a schematic diagram of the structure of the distributed control architecture of a large-scale energy storage power station included in the present invention.

[0052] Figure 2 It is a curve diagram of the command power and the actual output power of the energy storage power station included in the present invention.

[0053] Figure 3 This is a schematic diagram of the structure of the Ethernet control automation technology slave station device included in the present invention.

[0054] Figure 4This is a result diagram of the Ethernet control automation technology slave output included in the present invention. DETAILED DESCRIPTION

[0055] In order to make the technical personnel in this technical field better understand the present method, the technical scheme in this application will be described below in conjunction with the accompanying drawings. Obviously, the described implementation examples are only part of the embodiments of this application, rather than all the embodiments. Based on this embodiment, all other embodiments obtained by those skilled in the art without making creative work should fall within the scope of protection of this application.

[0056] The large-scale energy storage power station in this example has an installed capacity of 25MW / 100MWh and includes 10 energy storage battery packs. Each battery pack contains 10 energy storage units, each of which consists of two batteries and is equipped with a PCS with a rated power of 500kW. The energy storage unit is the control object, and its specific parameters are shown in Table 1. The maximum charge and discharge power is set to 1 / 4 of the rated capacity of the energy storage unit. Based on the differences in the storage energy of the energy storage unit, the initial SOC values ​​of the 100 energy storage units are randomly set in the range of 0.15-0.85.

[0057] Table 1 Energy storage unit parameters

[0058]

[0059] Step S1: Build a large-scale energy storage power station distributed control architecture: Figure 1 As shown, the distributed control architecture of the large-scale energy storage power station corresponding to this example is divided into two layers: the unit control layer and the information collection layer. The unit control layer contains 10 energy storage terminals, each of which controls one energy storage battery group. Among these energy storage terminals, there is a dominant energy storage terminal, which receives the dispatching instructions from the superior, starts the distributed algorithm of energy storage frequency modulation, and sends out an iteration signal. Other subordinate energy storage terminals exchange information according to the iteration signal. If they receive the end signal from the dominant energy storage terminal, they stop interacting and send the output signal of the controlled energy storage unit according to the iteration result. The information collection layer contains 10 battery groups, each of which is collected and controlled by the corresponding energy storage terminal, and each battery group contains 10 energy storage units.

[0060] Step S2: Design a distributed frequency modulation control strategy for the energy storage power station: Based on the first-order discrete weighted consensus algorithm, the SOC of each energy storage unit in the energy storage power station is balanced in the process of power optimization allocation. The corresponding expression is:

[0061]

[0062] Where: x is the state variable; k is the discrete time; H is the iteration matrix of the state variable; I n For n-order units

[0063] matrix, n is the total number of energy storage terminals, that is, 10; ε is the set iteration step; L is the Laplace matrix; W is the weighted diagonal matrix based on the battery state of charge SOC value, and the specific expression is shown in formula (2):

[0064] W=diag(w1,..., w i ,...,w n ) (2)

[0065] w i The specific expressions of are shown in equations (3) and (4), where equation (4) is used to obtain the relative weight of the battery state of charge SOC value:

[0066]

[0067] Where: SOC i is the SOC value currently collected by energy storage unit i; SOC max , SOC min They are the minimum and maximum values ​​of the SOC of the energy storage unit respectively.

[0068] As formula (1) is continuously iterated, the state variable x will gradually approach the weighted average of the initial value x0, where the expression of the initial value x0 is shown in formula (5):

[0069]

[0070] Where: P0 is an n×1 column matrix; P com It is the active power scheduling value sent by the superior to the dominant node 1.

[0071] Taking power as the state variable, the distributed computing of the energy storage power station is realized by combining formula (1) with the power P output equation. The specific expression is shown in formula (6):

[0072]

[0073] Using the idea of ​​block matrix, the constant term in the formula is converted into a block matrix: Assume that the SOC of each energy storage terminal is an n*1 column matrix. To meet the dimensional requirements of matrix subtraction, SOC max , SOC min Both are set to n*1 column matrices, and the former matrix elements are all SOC max , the latter matrix elements are all SOC min The division of formula (4) requires replacing the denominator with At this time, in order to ensure the consistency of the expression of the weight matrix W, W is divided into blocks, each w i is a diagonal matrix, and its specific expression is shown in formula (7):

[0074]

[0075] Similarly, I in formula (1) n L needs to be divided into blocks, and each element in the matrix is ​​converted into a block matrix. n For example, the expression after block division is shown in formula (8):

[0076]

[0077] In formula (5), P0 needs to be changed to n 2 ×1 column matrix, where P com Synchronously convert to an n×1 column matrix. Considering P com Each element of needs to be assigned a value, and the proportional distribution is considered according to the weighted matrix. The specific calculation formula is shown in formula (9):

[0078]

[0079] By partitioning each matrix into blocks, the design of a first-order discrete weighted consistency algorithm suitable for a hierarchical framework is achieved.

[0080] like Figure 2 As shown, compared with the centralized control, the distributed control method proposed in the present invention can make the actual output power of the energy storage power station follow the command power better. No matter the command power is too large or too small, the energy storage power station can be controlled to output the corresponding power.

[0081] Step S3: Build the Ethernet control automation technology master-slave communication system of the energy storage power station: In a large-scale energy storage power station, the Ethernet control automation technology master station is the energy storage terminal, and the slave station is the controller of the energy storage unit. The master station sends the calculated information such as the output power of the energy storage unit to the slave station through the Ethernet control automation technology data frame, so as to apply the Ethernet control automation technology to the large-scale energy storage power station. This example uses the Ethernet control automation technology master station, which is an Ethernet control automation technology controller written in pure rust language.

[0082] When receiving instructions from the upper layer, the Ethernet control automation technology master station needs to complete the initialization operation of the slave station, analyze the instructions, write the data frame content according to the requirements, and send it to the controlled slave stations. In addition, the master station needs to analyze and process the obtained slave station data and display it visually for the reference of developers. At the same time, the master station also needs to upload the data required by the upper layer.

[0083] Ethernet control automation technology slave devices need to communicate with the master station, and also need to control the application, output the data of the master station to the application, and accept the input data of the application and send it to the master station. Its structure is as follows Figure 3As shown. The slave controller is used to process and analyze Ethernet control automation technology data frames. When the data frame passes through the slave station, the slave controller will extract the control command sent to itself from it. After parsing the command, it will store the data to be output in the storage area and write the input data into the corresponding sub-message. The slave control microprocessor will read the data in the slave controller storage area and output it to the corresponding device. At the same time, the slave control microprocessor samples the feedback data from the device and sends the feedback data back to the slave controller as input data. The slave module used in this example is mainly composed of Beckhoff's EK1100 coupler and EL series products.

[0084] Step S4: Design and implement the Ethernet control automation technology program of the energy storage power station, including the following steps:

[0085] Step S401: Setting a maximum waiting time to prevent the program from being stuck due to communication failure. If the actual waiting time exceeds the set value, the loop is forced to exit and an error is reported to the user.

[0086] Step S402: In order to manage and coordinate the input and output of the slave station, the slave station sets a client array whose size is equal to the total number of slave stations, and each array element contains relevant information of a slave station. By calling the array, the input and output format of the slave station can be determined, so as to perform operations such as amplitude;

[0087] Step S403: After the master station energy storage terminal obtains the power allocation result through distributed computing, it formulates the operation cycle plan for data exchange between the master station and the slave station based on the polling cycle of the master station. If the data is sent in a cycle, it will cause unnecessary communication pressure. Therefore, the event-driven triggering method is used to trigger the data sending. The energy storage information transmitted by the slave station energy storage terminal is polled to ensure the real-time acquisition of the operation status of each energy storage unit.

[0088] To verify the rapidity of Ethernet control automation technology communication, the polling cycle remains at 1ms, but the number of slaves is increased. The output values ​​of all output slaves are increased by 1 in each polling cycle. The results are as follows: Figure 4 Although the number of slave devices increases, the number of messages in the master station does not increase, only the length of the data frame increases, so the communication time will not increase significantly, which ensures the real-time and fast communication of Ethernet control automation technology.

Claims

1. A distributed control method for a large-scale energy storage power station, characterized in that: The following steps are involved: Step S1: Building a distributed control architecture for a large-scale energy storage power station, the architecture includes a unit control layer and an information collection layer, wherein the unit control layer is responsible for local control and coordination of energy storage units in the energy storage power station, including at least one leading energy storage terminal and multiple subordinate energy storage terminals, and the information collection layer is responsible for real-time data collection and transmission; Step S2: Design a distributed frequency regulation control strategy for the energy storage power station. The strategy is based on a first-order discrete weighted consensus algorithm and uses a block matrix technology to convert the constant term in the algorithm into a block matrix. By dividing each matrix into blocks, the consistency algorithm can adapt to the design of a first-order discrete weighted consensus algorithm in a hierarchical framework. Step S3: Build an Ethernet control automation technology master-slave communication system for the energy storage power station, where the Ethernet control automation technology master station is the dominant energy storage terminal, responsible for calculating and outputting the power data information of the energy storage unit; the Ethernet control automation technology slave station is the subordinate energy storage terminal, responsible for receiving the data frames sent by the master station and executing the corresponding control instructions. The master station and the slave station achieve high-speed real-time communication through the Ethernet control automation technology protocol. Step S4: Design and implement the Ethernet control automation technology program of the energy storage power station, including: setting the maximum waiting time for communication and monitoring the communication delay, establishing a client array to manage the communication of multiple energy storage units, formulating the data exchange cycle between the Ethernet control automation technology master station and the slave station, and dynamically adjusting the communication frequency and data processing strategy according to actual needs.

2. A distributed control method for a large-scale energy storage power station according to claim 1, characterized in that: In step S1, the unit control layer includes multiple energy storage terminals, at least one of which is a dominant energy storage terminal, responsible for receiving superior scheduling instructions, starting the corresponding distributed algorithm, and sending iteration signals to other energy storage terminals; the unit control layer also includes multiple slave energy storage terminals connected to the energy storage unit, responsible for data collection and control of the energy storage unit, and information exchange according to the received iteration signals.

3. A distributed control method for a large-scale energy storage power station according to claim 1, characterized in that: In step S1, if the slave energy storage terminal receives an end signal from the master energy storage terminal, the information interaction is stopped, and a control signal is generated according to the final iteration result to adjust the output power of the controlled energy storage unit.

4. A distributed control method for a large-scale energy storage power station according to claim 1, characterized in that: In step S1, the information acquisition layer is responsible for connecting multiple battery packs, each battery pack is composed of multiple energy storage units through corresponding energy storage terminals, and the information acquisition layer collects and communicates data of each battery pack through each energy storage terminal.

5. A distributed control method for a large-scale energy storage power station according to claim 1, characterized in that: Step S2 includes the following steps: Step S201: construct a distributed frequency regulation control strategy model for a large-scale energy storage power station, which is expressed as shown in formula (1): Where: x is the state variable; k is the discrete time; H is the iteration matrix of the state variable; I n is the n-order unit matrix, n is the total number of energy storage terminals; ε is the set iteration step; L is the Laplace matrix; W is the weighted diagonal matrix based on the battery state of charge SOC value, and the specific expression is shown in formula (2): In=diag(in1,..., in i ,...,In n ) (2) Where: w i The calculation formulas are shown in equations (3) and (4), where equation (4) is used to obtain the relative weight of the battery state of charge SOC value: Where: SOC i is the battery state of charge SOC value currently collected by energy storage unit i; SOC max , SOC min They are the minimum and maximum values ​​of the state of charge (SOC) of the energy storage unit battery respectively; Step S202: Calculate the initial value x0 of the state variable x, as shown in formula (5): Where: P0 is an n×1 column matrix; P com is the active power dispatch value sent by the superior to the dominant node 1; Step S203: Using power as a state variable, combined with formula (1), the output power of the energy storage power station is subjected to distributed calculation, and the expression is shown in formula (6): Step S204: Based on the block matrix method, the constant term contained in equation (6) is converted into a block matrix: The SOC of formula (3) i Replace with the corresponding matrix elements and change SOC max and SOC min Equivalence conversion; Replace the denominator of the division in equation (4) with the corresponding matrix element and divide W into blocks so that each w i is a diagonal matrix; In formula (1), I n Block L and convert each element in the matrix into a block matrix; Replace P0 in equation (5) with the corresponding matrix element, and replace P com Equivalence conversion; Step S205: Considering the distribution problem between power and each energy storage unit, the weight coefficient of each energy storage unit is adjusted. The specific calculation formula is shown in formula (7):

6. A distributed control method for a large-scale energy storage power station according to claim 1, characterized in that: In step S3, the Ethernet control automation technology master station needs to perform the following operations: Initialization operation: The master station completes the initialization operation of each slave station according to the upper layer instructions; Instruction analysis and data frame generation: The master station analyzes the instruction content according to the received instruction, writes the corresponding data frame, and sends it to each controlled slave station; Data analysis and visualization: The master station analyzes and processes the data returned by the slave station and displays the results in a visual way for developers’ reference; Data upload: The master station uploads the data required by the superior according to the demand.

7. A distributed control method for a large-scale energy storage power station according to claim 1, characterized in that: In step S3, the Ethernet control automation technology slave device needs to perform the following operations: Data communication: bidirectional communication with the master station, outputting the data output by the master station to the application, and receiving the input data from the application and sending it back to the master station; Data frame processing and analysis: Processing and analyzing Ethernet control automation technology data frames received from the master station; Control device operation and data sampling: The slave control microprocessor reads the data in the slave controller storage area and outputs it to the corresponding device; at the same time, the slave control microprocessor samples the feedback data from the device and sends the feedback data back to the slave controller storage area as input data for storage.

8. A distributed control method for a large-scale energy storage power station according to claim 1, characterized in that: Step S4 includes the following steps: Step S401: Set a maximum waiting time. If the actual waiting time exceeds the set value, the loop is forced to exit and an error is reported to the user. Step S402: The slave station sets a client array whose size is equal to the total number of slave stations, and each array element contains relevant information of a slave station. By calling the array, the input and output format of the slave station can be determined, so as to perform operations such as amplitude; Step S403: After the master station energy storage terminal obtains the power allocation result through distributed computing, it formulates the operation cycle plan for data exchange between the master station and the slave station based on the polling cycle of the master station, and uses event-driven triggering to trigger data distribution. The energy storage information transmitted by the slave station energy storage terminal is polled.

Images