Coordinated control method and system for distribution network with distributed photovoltaic storage

Abstract

A coordinated control method and system for a distribution network with distributed photovoltaic storage is proposed. According to the gradient and standard deviation of the net load power in each period of the day-ahead scheduling plan stage, the direction change frequency and standard deviation, the net load power curve is divided into different resolution periods; the day-ahead scheduling model is established with the minimum total cost of the distribution network as the optimization goal; the population fitness is dynamically adjusted in different resolution periods to improve the sparrow optimization algorithm, the day-ahead scheduling model is solved by the improved sparrow optimization algorithm, and the day-ahead scheduling plan is determined by the entropy weight method-TOPSIS method; the intraday scheduling time period is determined based on each rolling optimization time scale; the intraday scheduling model is established with the minimum difference between the output power forecast value of the photovoltaic power station in the intraday scheduling time period and the output power plan value of the photovoltaic power station in the day-ahead scheduling plan as the optimization goal; the optimal solution of the intraday scheduling model is used as the intraday scheduling plan. Flexible coordinated control of the distribution network is realized.

Description

Technical Field

[0001] The present invention belongs to the technical field of distribution network optimization control, and in particular relates to a distribution network coordinated control method and system containing distributed photovoltaic storage. Background Art

[0002] With the large-scale development and utilization of distributed energy, especially distributed photovoltaics, and the decline in the construction and operation costs of energy storage, distributed photovoltaic storage has received extensive attention in recent years. In view of the many advantages of distributed photovoltaic storage and the continuous affirmation of the dominant position of energy storage, the large-scale development of distributed photovoltaic storage is the only way forward.

[0003] In the existing technology, the research on energy management of distribution network mainly focuses on the design of the topological structure of hybrid distribution network and the realization of economic distribution of output power of each controllable distributed unit in the distribution network to ensure the stability of distribution network operation. In the new power system, a large number of distributed power sources are connected to the distribution network, but the relevant control and regulation means are lacking.

[0004] Considering the multi-time scale characteristics of intermittent renewable energy output forecast and load forecast, that is, the predicted values ​​of active power are different at different time scales. Therefore, the day-ahead dispatch plan with a single time scale cannot well reflect the forecast error of intermittent distributed power output and load demand and the impact of unplanned instantaneous power fluctuations on the optimal operation of AC / DC hybrid distribution network, and does not conform to the actual operation of the distribution network. Summary of the invention

[0005] In order to solve the deficiencies in the prior art, the present invention provides a coordinated control method and system for a distribution network containing distributed photovoltaic storage, which realizes the coordinated interaction between the photovoltaic storage system and the controllable load and the distribution network, and helps to significantly improve the flexibility of the distribution network. It improves the economic efficiency and stability of the distribution network operation, promotes photovoltaic consumption, and ensures the safe, economical and stable operation of the distribution network.

[0006] The present invention adopts the following technical solution.

[0007] The present invention proposes a coordinated control method for a distribution network containing distributed photovoltaic storage. The coordinated control includes: a day-ahead scheduling plan stage and an intraday scheduling plan stage, and respectively formulates a day-ahead scheduling plan and an intraday scheduling plan for a hybrid distribution network containing distributed photovoltaic storage, including: Obtain the total load power and the output power of the photovoltaic power station in the distribution network to determine the net load power curve of the distribution network; divide the net load power curve into different resolution time periods according to the gradient and standard deviation, direction change frequency and standard deviation of the net load power in each time period during the day-ahead scheduling planning stage; Taking the minimization of the total cost of distribution network operation as the optimization goal, a day-ahead dispatching model is established; the fitness of the population is dynamically adjusted in different resolution periods to improve the sparrow optimization algorithm, and the improved sparrow optimization algorithm is used to solve the day-ahead dispatching model. For the optimal solution obtained, the entropy weight method-TOPSIS method is used to determine the day-ahead dispatching plan; The maximum value of the time scale that satisfies different resolution periods is taken as the rolling optimization time scale, and the intraday scheduling time period is determined based on each rolling optimization time scale; the difference between the predicted output power of the photovoltaic power station within the intraday scheduling time period and the planned output power of the photovoltaic power station in the day-ahead scheduling plan is minimized as the optimization goal, and an intraday scheduling model is established; the optimal solution of the intraday scheduling model is used as the intraday scheduling plan.

[0008] Preferably, the net load power of the distribution network is a variable time period data set that satisfies the following relationship: , In the formula, for Net load power at the moment, , They are The total load power and the output power of the photovoltaic power station at the moment.

[0009] Preferably, the net load power is calculated in the time period The gradient inside and standard deviation , Direction change frequency Sum and Standard Deviation ; Set the net load power in the time period The mean gradient within and standard deviation , the average frequency of direction changes With standard deviation In order to divide the thresholds of the variable time period data set, the variable time period data set is divided into three categories using the set four thresholds, including: 1) Satisfy both , , , When the time period is defined as the first resolution time period , the first resolution period is based on a time scale of 15 min; 2) Satisfy both , , , The time period is defined as the second resolution period , the second resolution period is based on 30 minutes; 3) Satisfy both , , , The time period is defined as the third resolution period , the third resolution period is based on a time scale of 60 minutes.

[0010] Preferably, the minimum total cost of distribution network operation is taken as the optimization goal, and a day-ahead dispatching model is established to satisfy the following relationship: , In the formula, is the total cost of distribution network operation, , They are the investment cost and operation and maintenance cost of the distribution network; Penalty cost for curtailment of electricity, The energy storage life loss cost.

[0011] Preferably, the life loss cost of a battery during one discharge satisfies the following relationship: , In the formula, is the life loss cost of a battery discharge, is the initial investment for the battery, is the state of charge of the battery, , and are fitting constants, The rated throughput of the battery is , The battery is discharged at rated current Rated capacity under is the rated depth of discharge, Rated discharge depth and rated discharge current Rated cycle life under .

[0012] Preferably, the life loss cost of a battery discharge is dynamically adjusted for different resolution time periods to obtain the energy storage life loss cost, which satisfies the following relationship: , In the formula, is the adjustment coefficient of the life loss cost of a battery discharge based on different resolution periods. , the second resolution period , the first resolution period .

[0013] Preferably, an improved sparrow optimization algorithm is used to solve the day-ahead scheduling model, including: 1) Read the photovoltaic day-ahead forecast data, load day-ahead forecast data and peak-valley electricity price data; 2) Initialize the population velocity and position to satisfy the following relationship: , In the formula, For particles The next Dimensional variables, , is the total number of particles, , is the total number of variable dimensions, is a random number between 0 and 1, For the initial particle The next Dimensional variables; 3) In different resolution periods, based on the day-ahead scheduling model, the population fitness is calculated and dynamically adjusted to satisfy the following relationship: , In the formula, and are the fitness of the population before and after dynamic adjustment, is the adjustment coefficient of the population fitness based on different resolution periods. , the second resolution period , the first resolution period ; 4) Take the optimal value from the dynamically adjusted fitness and use the sparrow as the discoverer and update the position. The rest of the sparrows are followers and update the position. Some sparrows are randomly selected from the followers as sentinels and update the position, so as to determine the optimal value of the individual and the optimal value of the group. 5) Based on the reverse learning method, after introducing disturbances to the optimal values ​​of the individual and the group, new optimal values ​​of the individual and the group are generated, satisfying the following relationship: , Where: For the t Individual optimal value and group optimal value of the generation population The new individual optimal value and group optimal value of the population generated after the disturbance is introduced. , are the upper and lower bounds of the decision variables, is a 1× The random number matrix, is the space dimension, is the XOR symbol; 6) Repeat step 3) to obtain the optimal value of the individual and the group Fitness , the new optimal value of the individual population and the optimal value of the group Fitness , select the optimal fitness value as the global optimal value of this iteration; 7) Introduce an external population to record Pareto non-dominated solutions, calculate the congestion degree, add new non-dominated solutions to the external population and remove some solutions from the external population; 8) Determine whether the termination conditions are met. If so, end the solution and output the optimal solution. For the optimal solution obtained, use the entropy weight method-TOPSIS method to determine the day-ahead scheduling plan.

[0014] Preferably, the day-ahead scheduling plan includes: a planned output power value of a photovoltaic power station and an energy storage charging and discharging scheduling plan.

[0015] Preferably, the time scales of different resolution time periods are obtained, and the greatest common divisor of the time scales is used as the rolling optimization time scale.

[0016] Preferably, determining the intraday scheduling plan includes: Based on each rolling optimization time scale, the intraday scheduling time period is determined, the actual output power value of the photovoltaic power station within the intraday scheduling time period and the historical output power value of the photovoltaic power station are obtained, and the long short-term memory network is used to output the predicted value of the output power increment of the photovoltaic power station within the intraday scheduling time period. The output power prediction value of the photovoltaic power station is determined by the following relationship: , In the formula, For Always The predicted output power of the photovoltaic power station at any time, is the scheduling time step, For The initial output power value of the photovoltaic power station at the moment, For Always The predicted output power increment of the photovoltaic power station at any time; Taking the minimum difference between the predicted output power of the photovoltaic power station within the daily scheduling time period and the planned output power of the photovoltaic power station in the day-ahead scheduling plan as the optimization goal, an intraday scheduling model is established; The intraday scheduling model satisfies the following relationship: , In the formula, The difference between the predicted output power of the photovoltaic power station and the planned output power of the photovoltaic power station in the day-ahead scheduling plan is the smallest. For Always The predicted output power of the photovoltaic power station at any time, For the day-ahead scheduling plan The planned output power value of the photovoltaic power station at the time, is an intraday scheduling time period, is the error fluctuation penalty factor of the photovoltaic power station output power; The actual output power value of the photovoltaic power station in the current intraday scheduling time period is obtained as the initial output power value of the photovoltaic power station in the next intraday scheduling time period.

[0017] Preferably, the MOSEK solver is used to solve the intraday scheduling model, and the optimal solution obtained is the intraday scheduling optimization value within each intraday scheduling time period, which constitutes the intraday scheduling plan.

[0018] The present invention also proposes a distribution network coordination control system including distributed photovoltaic storage, comprising: The data processing module is used to obtain the total load power and the output power of the photovoltaic power station in the distribution network to determine the net load power curve of the distribution network; the net load power curve is divided into different resolution time periods according to the gradient and standard deviation, direction change frequency and standard deviation of the net load power in each time period in the day-ahead scheduling planning stage; The day-ahead dispatch planning module is used to establish a day-ahead dispatch model with the minimum total cost of distribution network operation as the optimization goal; the fitness of the population is dynamically adjusted in different resolution periods to improve the sparrow optimization algorithm, and the improved sparrow optimization algorithm is used to solve the day-ahead dispatch model. For the optimal solution obtained, the entropy weight method-TOPSIS method is used to determine the day-ahead dispatch plan; The intraday scheduling module is used to determine the intraday scheduling time period based on the rolling optimization time scale, taking the maximum value of the time scale that satisfies different resolution periods at the same time as the rolling optimization time scale; the intraday scheduling model is established with the minimum difference between the predicted output power value of the photovoltaic power station within the intraday scheduling time period and the planned output power value of the photovoltaic power station in the day-ahead scheduling plan as the optimization target; and the optimal solution of the intraday scheduling model is used as the intraday scheduling plan.

[0019] The present invention also provides a terminal, comprising a processor and a storage medium; the storage medium is used to store instructions; the processor is used to operate according to the instructions to execute the steps of the method.

[0020] The present invention also provides a computer-readable storage medium on which a computer program is stored, and the program implements the steps of the method when executed by a processor.

[0021] The beneficial effects of the present invention are that, compared with the prior art, it at least includes utilizing the multi-time-scale energy collaborative optimization control technology of the distribution network containing distributed photovoltaic storage, combining power forecasting, day-ahead scheduling method, intraday rolling optimization method and hierarchical and partitioned control strategy, and designing a distribution network energy management system containing distributed photovoltaic storage, realizing real-time balance between each power generation unit and the demand-side load and joint optimization operation of multiple energy systems, improving the on-site absorption capacity of intermittent renewable energy, and thus improving the comprehensive energy utilization efficiency of the distribution network containing distributed photovoltaic storage. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] Figure 1 It is a flow chart of a coordinated control method for a distribution network including distributed photovoltaic storage. DETAILED DESCRIPTION

[0023] In order to make the purpose, technical scheme and advantages of the present invention clearer, the technical scheme of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. The embodiments described in this application are only embodiments of a part of the present invention, rather than all embodiments. Based on the spirit of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work belong to the protection scope of the present invention.

[0024] The present invention proposes a coordinated control method for a distribution network containing distributed photovoltaic storage. The coordinated control includes: a day-ahead scheduling planning stage and an intraday scheduling planning stage. The coordinated control method for a distribution network containing distributed photovoltaic storage starts from the two scheduling stages of day-ahead and intraday, combines the concept of hierarchical zoning, and formulates a day-ahead scheduling plan and an intraday scheduling plan for a hybrid distribution network containing distributed photovoltaic storage respectively.

[0025] like Figure 1 As shown, the coordinated control method includes: Step 1: Obtain the total load power and the output power of the photovoltaic power station in the distribution network to determine the net load power curve of the distribution network; divide the net load power curve into different resolution time periods according to the gradient and standard deviation, direction change frequency and standard deviation of the net load power in each time period in the day-ahead scheduling planning stage.

[0026] Specifically, the net load power of the distribution network is a variable time period data set that satisfies the following relationship: , In the formula, for Net load power at the moment, , They are The total load power and the output power of the photovoltaic power station at the moment.

[0027] In order to analyze the energy storage output plan and establish an economic dispatch model containing a distributed photovoltaic storage distribution network, the present invention introduces the output of the photovoltaic power station into the original load demand to obtain the net load demand. After a high proportion of new energy is incorporated into the system, its net load demand fluctuates more violently and frequently than the original load. The increase in uncertainty makes it impossible to accurately track the net load by relying solely on thermal power units, so additional energy storage is selected to assist in dispatching.

[0028] Calculate the net load power in the time period The gradient inside and standard deviation , Direction change frequency Sum and Standard Deviation ; Set the net load power in the time period The mean gradient within and standard deviation , the average frequency of direction changes With standard deviation , is the threshold for dividing the variable time period data set. In order to reduce the complexity of the model, the variable time period data set is divided into three categories using the set four thresholds, including: 1) Satisfy both , , , When the net load power is in the period During this period, the amplitude fluctuates violently, and the gradient and direction change frequencies are high. Therefore, this period is defined as the first resolution period. ,The first resolution period is based on 15 minutes time scale to cope with the more complex ,net load characteristic curve.

[0029] 2) Satisfy both , , , When the net load power is in the period During this period, the amplitude fluctuates violently, but the gradient and direction change frequencies are low. Therefore, this period is defined as the second resolution period. The second resolution period is based on a time scale of 30 minutes to cope with the net load characteristic curve with an obvious rising and falling trend.

[0030] 3) Satisfy both , , , When the net load power is in the period Within this period, the amplitude does not fluctuate violently, and the net load characteristic curve generally presents a stable trend. Therefore, this period is defined as the third resolution period. , the third resolution period is based on a time scale of 60 minutes.

[0031] Step 2: Taking the minimization of the total cost of distribution network operation as the optimization goal, a day-ahead dispatching model is established; the fitness of the population is dynamically adjusted in different resolution periods to improve the sparrow optimization algorithm, and the improved sparrow optimization algorithm is used to solve the day-ahead dispatching model. For the optimal solution obtained, the entropy weight method-TOPSIS method is used to determine the day-ahead dispatching plan.

[0032] Day-ahead optimization scheduling is to obtain photovoltaic output forecast data, controllable distributed power supply operation data, energy storage equipment operation data, adjustable load operation data and non-adjustable load forecast data to generate a day-ahead scheduling plan for the distribution network; the day-ahead scheduling plan is to formulate an operation plan for the next day based on the load and renewable energy generation forecast curve, which needs to comprehensively consider the system economic operation and reliability plan. Battery energy storage has an important impact on the formulation of the system plan.

[0033] Considering the differences in control strategies under different resolution periods, and considering that the differences in control strategies cannot reduce the generalization ability of the day-ahead dispatch model, the day-ahead dispatch model is established with the minimum total cost of distribution network operation as the optimization goal, satisfying the following relationship: , In the formula, is the total cost of distribution network operation, , They are the investment cost and operation and maintenance cost of the distribution network; Penalty cost for curtailment of electricity, The energy storage life loss cost.

[0034] When the distribution network system dispatches battery energy storage to smooth out fluctuations in renewable energy or to reduce peak loads and fill valleys, the life of the energy storage itself will be lost. The life loss cost of a battery discharge satisfies the following relationship: , In the formula, is the life loss cost of a battery discharge, is the initial investment for the battery, is the state of charge of the battery, , and are fitting constants, The rated throughput of the battery is , The battery is discharged at rated current Rated capacity under is the rated depth of discharge, Rated discharge depth and rated discharge current Rated cycle life under .

[0035] In different resolution periods, since the intensity and frequency of net load power fluctuations are different, the strength of energy storage auxiliary scheduling is also different. Taking the life loss cost of a battery discharge in the third resolution period as a benchmark, the frequent energy storage charging and discharging in the first resolution period and the second resolution period both cause an increase in the battery life loss cost. Therefore, the present invention proposes to dynamically adjust the life loss cost of a battery discharge in different resolution periods, thereby realizing dynamic adjustment of the total cost of distribution network operation, satisfying the following relationship: , In the formula, is the adjustment coefficient of the life loss cost of a battery discharge based on different resolution periods. , the second resolution period , the first resolution period .

[0036] In the embodiment, the regulation coefficient is determined by performing charge and discharge tests of different frequencies on the battery. .

[0037] Since the time scales in different resolution periods are different, and in order to accurately track the net load and maintain the consistency of the day-ahead scheduling strategy, the present invention proposes to solve the day-ahead scheduling model by using an improved sparrow optimization algorithm, dynamically adjust the fitness of the population for different resolution periods, and improve the accurate tracking characteristics of the optimal value for the net load power when determining the individual optimal value and the group optimal value of the population. The improved sparrow optimization algorithm is used to solve the day-ahead scheduling model, including: 1) Read the photovoltaic day-ahead forecast data, load day-ahead forecast data, and peak and valley electricity price data.

[0038] 2) Initialize the population speed and position. Use the tent chaos initialization strategy to generate a uniformly distributed and sufficient initialization population in the search space for boundary constraints.

[0039] , In the formula, For particles The next Dimensional variables, , is the total number of particles, , is the total number of variable dimensions, is a random number between 0 and 1, For the initial particle The next Dimensional variables.

[0040] 3) In different resolution periods, based on the day-ahead scheduling model, the population fitness is calculated and dynamically adjusted to satisfy the following relationship: , In the formula, and are the fitness of the population before and after dynamic adjustment, is the adjustment coefficient of the population fitness based on different resolution periods. , the second resolution period , the first resolution period After the adjustment coefficient is introduced, in the first resolution period, the excellent individuals in the sparrow population with better fitness values ​​are retained as much as possible, and the diversity of the population is increased in the third resolution period, thereby effectively improving the accurate tracking of the net load power by the optimal value and taking into account the search ability of the algorithm.

[0041] 4) Take the optimal value from the dynamically adjusted fitness and use the sparrow as the discoverer and update the position. The rest of the sparrows are followers and update their positions. Some sparrows are randomly selected from the followers as sentinels and update their positions, so as to determine the individual optimal value and the group optimal value of the population.

[0042] 5) Based on the reverse learning method, after introducing disturbances to the optimal values ​​of the individual and the group, new optimal values ​​of the individual and the group are generated, satisfying the following relationship: , Where: For the t Individual optimal value and group optimal value of the generation population The new individual optimal value and group optimal value of the population generated after the disturbance is introduced. , are the upper and lower bounds of the decision variables, is a 1× The random number matrix, is the space dimension, Is the XOR symbol.

[0043] 6) Repeat step 3) to obtain the optimal value of the individual and the group Fitness , the new optimal value of the individual population and the optimal value of the group Fitness , select the optimal fitness value as the global optimal value of this iteration.

[0044] In the process of solving the day-ahead scheduling model by using the improved sparrow optimization algorithm, the present invention also adopts a reverse learning method to generate new optimal values ​​of the individual population and the group, and then compares the optimal values ​​of the individual population and the group. Fitness , the new optimal value of the individual population and the optimal value of the group Fitness , taking the optimal fitness value as the global optimal value of this iteration, so that the individual sparrow can better find the optimal solution.

[0045] 7) Introduce an external population to record Pareto non-dominated solutions, calculate the congestion degree, add new non-dominated solutions to the external population and remove some solutions from the external population.

[0046] 8) Determine whether the termination conditions are met. If so, end the solution and output the optimal solution. For the optimal solution obtained, use the entropy weight method-TOPSIS method to determine the day-ahead scheduling plan.

[0047] In the embodiment, the day-ahead scheduling plan includes, but is not limited to: a planned output power value of a photovoltaic power station and an energy storage charging and discharging scheduling plan.

[0048] Step 3, taking the maximum value of the time scale that satisfies different resolution time periods as the rolling optimization time scale, and determining the intraday scheduling time period based on each rolling optimization time scale; taking the minimum difference between the output power forecast value of the photovoltaic power station within the intraday scheduling time period and the output power plan value of the photovoltaic power station in the day-ahead scheduling plan as the optimization goal, establishing an intraday scheduling model; and taking the optimal solution of the intraday scheduling model as the intraday scheduling plan.

[0049] Intraday optimization scheduling is to reduce the power fluctuation of the system by referring to the day-ahead scheduling plan and real-time updated light information during the intraday stage, reducing the predicted output error of photovoltaic power generation, formulating a rolling limited-time domain optimization scheduling plan, and outputting the optimization scheduling results.

[0050] Obtain the time scales of different resolution periods, and use the greatest common divisor of each time scale as the rolling optimization time scale. In this example, 15 minutes is used as the rolling optimization time scale. Through rolling optimization, an intraday scheduling plan is formulated. Based on the intraday scheduling plan at the current moment and the predicted feedback on the system's operating status at the next moment, the scheduling plan for the next rolling optimization moment is obtained. During the day-ahead optimization scheduling, the photovoltaic output is affected by the light intensity, and there is a large prediction error. The output of each device in the distribution network fluctuates greatly. Shortening the prediction time can effectively reduce the photovoltaic output prediction error. Intraday optimization scheduling uses the day-ahead optimization scheduling results as a reference, and performs multi-time scale optimization scheduling within a preset time step.

[0051] Specifically, determining the intraday scheduling plan includes: Step 3.1, based on each rolling optimization time scale, determine the intraday scheduling time period, obtain the actual value of the photovoltaic power station output power within the intraday scheduling time period and the historical value of the photovoltaic power station output power, use the long short-term memory network to output the predicted value of the photovoltaic power station output power increment within the intraday scheduling time period, and determine the predicted value of the photovoltaic power station output power by the following relationship: , In the formula, For Always The predicted output power of the photovoltaic power station at any time, is the scheduling time step, For The initial output power value of the photovoltaic power station at the moment, For Always The predicted output power increment of the photovoltaic power station at any moment.

[0052] In the embodiment, The time is the starting time of a daily scheduling time cycle.

[0053] Step 3.2, taking the minimum difference between the predicted output power of the photovoltaic power station within the daily scheduling time period and the planned output power of the photovoltaic power station in the day-ahead scheduling plan as the optimization goal, establishing an intraday scheduling model; The intraday optimization scheduling target based on MPC is to use the optimization scheduling result of photovoltaic power generation on the day before as the benchmark, impose a penalty on the difference of photovoltaic power generation output in adjacent time periods, so as to minimize the difference between the photovoltaic active power generation output and the benchmark value at each moment of the day, thereby reducing the predicted photovoltaic power generation error and reducing the power fluctuation of the system.

[0054] The intraday scheduling model satisfies the following relationship: , In the formula, The difference between the predicted output power of the photovoltaic power station and the planned output power of the photovoltaic power station in the day-ahead scheduling plan is the smallest. For Always The predicted output power of the photovoltaic power station at any time, For the day-ahead scheduling plan The planned output power value of the photovoltaic power station at the time, is an intraday scheduling time period, is the error fluctuation penalty factor of the output power of the photovoltaic power station.

[0055] Step 3.3, obtaining the actual value of the photovoltaic power station output power in the current intraday scheduling time period as the initial value of the photovoltaic power station output power in the next intraday scheduling time period; In the embodiment, the initial value of the output power of the photovoltaic power station is obtained, and the initial time of the intraday optimization scheduling is set The prediction time period of the photovoltaic power station output power is 15min×3, and the intraday scheduling time period is 15min×4. The historical value of the photovoltaic power station output power is obtained in the first 15min×3 period of the intraday scheduling time period, and the actual value of the photovoltaic power station output power is obtained in the last 15min period as the initial value of the photovoltaic power station output power in the next intraday scheduling time period. .

[0056] The intraday rolling optimization scheduling based on MPC adds a feedback correction link on the basis of the prediction model. According to the predicted value of the day-ahead optimization scheduling and the actual measured value at the current moment, the initial value of the photovoltaic output of the next cycle of the prediction model is corrected, thereby reducing the error caused by the rolling prediction.

[0057] In the embodiment, the MOSEK solver is used to solve the intraday scheduling model, and the optimal solution obtained is the intraday scheduling optimization value within each intraday scheduling time period, which constitutes the intraday scheduling plan.

[0058] The present invention also proposes a distribution network coordination control system including distributed photovoltaic storage, comprising: The data processing module is used to obtain the total load power and the output power of the photovoltaic power station in the distribution network to determine the net load power curve of the distribution network; the net load power curve is divided into different resolution time periods according to the gradient and standard deviation, direction change frequency and standard deviation of the net load power in each time period in the day-ahead scheduling planning stage; The day-ahead dispatch planning module is used to establish a day-ahead dispatch model with the minimum total cost of distribution network operation as the optimization goal; the fitness of the population is dynamically adjusted in different resolution periods to improve the sparrow optimization algorithm, and the improved sparrow optimization algorithm is used to solve the day-ahead dispatch model. For the optimal solution obtained, the entropy weight method-TOPSIS method is used to determine the day-ahead dispatch plan; The intraday scheduling module is used to determine the intraday scheduling time period based on the rolling optimization time scale, taking the maximum value of the time scale that satisfies different resolution periods at the same time as the rolling optimization time scale; the intraday scheduling model is established with the minimum difference between the predicted output power value of the photovoltaic power station within the intraday scheduling time period and the planned output power value of the photovoltaic power station in the day-ahead scheduling plan as the optimization target; and the optimal solution of the intraday scheduling model is used as the intraday scheduling plan.

[0059] The present disclosure may be a system, a method and / or a computer program product. The computer program product may include a computer-readable storage medium carrying computer-readable program instructions for causing a processor to implement various aspects of the present disclosure.

[0060] A computer-readable storage medium may be a tangible device that can hold and store instructions used by an instruction execution device. A computer-readable storage medium may be, for example, but not limited to, an electrical storage device, a magnetic storage device, an optical storage device, an electromagnetic storage device, a semiconductor storage device, or any suitable combination of the above. More specific examples (a non-exhaustive list) of computer-readable storage media include: a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), a static random access memory (SRAM), a portable compact disk read-only memory (CD-ROM), a digital versatile disk (DVD), a memory stick, a floppy disk, a mechanical encoding device, such as a punch card or a raised structure in a groove on which instructions are stored, and any suitable combination of the above. The computer-readable storage medium used herein is not to be interpreted as a transient signal itself, such as a radio wave or other freely propagating electromagnetic wave, an electromagnetic wave propagating through a waveguide or other transmission medium (e.g., a light pulse through a fiber optic cable), or an electrical signal transmitted through a wire.

[0061] The computer-readable program instructions described herein can be downloaded from a computer-readable storage medium to each computing / processing device, or downloaded to an external computer or external storage device via a network, such as the Internet, a local area network, a wide area network, and / or a wireless network. The network can include copper transmission cables, optical fiber transmissions, wireless transmissions, routers, firewalls, switches, gateway computers, and / or edge servers. The network adapter card or network interface in each computing / processing device receives the computer-readable program instructions from the network and forwards the computer-readable program instructions for storage in the computer-readable storage medium in each computing / processing device.

[0062] The computer program instructions for performing the operations of the present disclosure may be assembly instructions, instruction set architecture (ISA) instructions, machine instructions, machine-dependent instructions, microcode, firmware instructions, state setting data, or source code or object code written in any combination of one or more programming languages, including object-oriented programming languages, such as Smalltalk, C++, etc., and conventional procedural programming languages, such as "C" language or similar programming languages. The computer-readable program instructions may be executed entirely on the user's computer, partially on the user's computer, as a separate software package, partially on the user's computer, partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., using an Internet service provider to connect through the Internet). In some embodiments, by using the state information of the computer-readable program instructions to personalize an electronic circuit, such as a programmable logic circuit, a field programmable gate array (FPGA), or a programmable logic array (PLA), the electronic circuit may execute the computer-readable program instructions, thereby implementing various aspects of the present disclosure.

[0063] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the above embodiments, ordinary technicians in the relevant field should understand that the specific implementation methods of the present invention can still be modified or replaced by equivalents, and any modifications or equivalent replacements that do not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.

Claims

1. A coordinated control method for a distribution network including distributed photovoltaic storage, the coordinated control comprising: In the day-ahead dispatch planning stage and the intraday dispatch planning stage, a day-ahead dispatch plan and an intraday dispatch plan of a hybrid distribution network including distributed photovoltaic storage are formulated respectively, which are characterized by including: Obtain the total load power and the output power of the photovoltaic power station in the distribution network to determine the net load power curve of the distribution network; divide the net load power curve into different resolution time periods according to the gradient and standard deviation, direction change frequency and standard deviation of the net load power in each time period during the day-ahead scheduling planning stage; Taking the minimization of the total cost of distribution network operation as the optimization goal, a day-ahead dispatching model is established; the fitness of the population is dynamically adjusted in different resolution periods to improve the sparrow optimization algorithm, and the improved sparrow optimization algorithm is used to solve the day-ahead dispatching model. For the optimal solution obtained, the entropy weight method-TOPSIS method is used to determine the day-ahead dispatching plan; The maximum value of the time scale that satisfies different resolution periods is taken as the rolling optimization time scale, and the intraday scheduling time period is determined based on each rolling optimization time scale; the difference between the predicted value of the photovoltaic power station output power within the intraday scheduling time period and the planned value of the photovoltaic power station output power in the day-ahead scheduling plan is minimized as the optimization goal, and an intraday scheduling model is established; the optimal solution of the intraday scheduling model is used as the intraday scheduling plan.

2. The coordinated control method for a distribution network containing distributed photovoltaic storage according to claim 1 is characterized in that: The net load power of the distribution network is a variable time period data set that satisfies the following relationship: ， In the formula, for Net load power at the moment, , They are The total load power and the output power of the photovoltaic power station at the moment.

3. The coordinated control method for a distribution network containing distributed photovoltaic storage according to claim 2 is characterized in that: Calculate the net load power in the time period The gradient inside and standard deviation , Direction change frequency Sum and Standard Deviation ; Set the net load power in the time period The mean gradient within and standard deviation , the average frequency of direction changes With standard deviation In order to divide the thresholds of the variable time period data set, the variable time period data set is divided into three categories using the set four thresholds, including: 1) Satisfy both , , , When the time period is defined as the first resolution time period , the first resolution period is based on a time scale of 15 min; 2) Satisfy both , , , The time period is defined as the second resolution period , the second resolution period is based on 30 minutes; 3) Satisfy both , , , The time period is defined as the third resolution period , the third resolution period is based on a time scale of 60 minutes.

4. The coordinated control method for a distribution network containing distributed photovoltaic storage according to claim 1 is characterized in that: Taking the minimum total cost of distribution network operation as the optimization goal, a day-ahead dispatching model is established to satisfy the following relationship: ， In the formula, is the total cost of distribution network operation, , They are the investment cost and operation and maintenance cost of the distribution network; Penalty cost for curtailment of electricity, The energy storage life loss cost.

5. The coordinated control method for a distribution network containing distributed photovoltaic storage according to claim 4 is characterized in that: The life loss cost of a battery discharge satisfies the following relationship: ， In the formula, is the life loss cost of a battery discharge, is the initial investment for the battery, is the state of charge of the battery, , and are fitting constants, The rated throughput of the battery is , The battery is discharged at rated current Rated capacity under is the rated depth of discharge, Rated discharge depth and rated discharge current Rated cycle life under .

6. The coordinated control method for a distribution network containing distributed photovoltaic storage according to claim 5 is characterized in that: The life loss cost of a battery discharge is dynamically adjusted for different resolution periods to obtain the energy storage life loss cost, which satisfies the following relationship: ， In the formula, is the adjustment coefficient of the life loss cost of a battery discharge based on different resolution periods. , the second resolution period , the first resolution period .

7. The coordinated control method for a distribution network containing distributed photovoltaic storage according to claim 1 is characterized in that: The improved sparrow optimization algorithm is used to solve the day-ahead scheduling model, including: 1) Read the photovoltaic day-ahead forecast data, load day-ahead forecast data and peak-valley electricity price data; 2) Initialize the population velocity and position to satisfy the following relationship: ， In the formula, For particles The next Dimensional variables, , is the total number of particles, , is the total number of variable dimensions, is a random number between 0 and 1, The initial particle The next Dimensional variables; 3) In different resolution periods, based on the day-ahead scheduling model, the population fitness is calculated and dynamically adjusted to satisfy the following relationship: ， In the formula, and are the fitness of the population before and after dynamic adjustment, is the adjustment coefficient of the population fitness based on different resolution periods. , the second resolution period , the first resolution period ; 4) Take the optimal value from the dynamically adjusted fitness and use the sparrow as the discoverer and update the position. The rest of the sparrows are followers and update the position. Some sparrows are randomly selected from the followers as sentinels and update the position, so as to determine the optimal value of the individual and the optimal value of the group. 5) Based on the reverse learning method, after introducing disturbances to the optimal values ​​of the individual and the group, new optimal values ​​of the individual and the group are generated, satisfying the following relationship: ， Where: For the t Individual optimal value and group optimal value of the generation population The new individual optimal value and group optimal value of the population generated after the disturbance is introduced. , are the upper and lower bounds of the decision variables, is a 1× The random number matrix, is the space dimension, is the XOR symbol; 6) Repeat step 3) to obtain the optimal value of the individual and the group Fitness , the new optimal value of the individual population and the optimal value of the group Fitness , select the optimal fitness value as the global optimal value of this iteration; 7) Introduce an external population to record Pareto non-dominated solutions, calculate the congestion degree, add new non-dominated solutions to the external population and remove some solutions from the external population; 8) Determine whether the termination conditions are met. If so, end the solution and output the optimal solution. For the optimal solution obtained, use the entropy weight method-TOPSIS method to determine the day-ahead scheduling plan.

8. The coordinated control method for a distribution network including distributed photovoltaic storage according to claim 7 is characterized in that: The day-ahead dispatch plan includes: the planned output power value of the photovoltaic power station and the energy storage charging and discharging dispatch plan.

9. The coordinated control method for a distribution network containing distributed photovoltaic storage according to claim 3 is characterized in that: The time scales of different resolution periods are obtained, and the greatest common divisor of each time scale is used as the rolling optimization time scale.

10. The coordinated control method for a distribution network with distributed photovoltaic storage according to claim 4 is characterized in that: Determine the intraday scheduling plan including: Based on each rolling optimization time scale, the intraday scheduling time period is determined, the actual output power value of the photovoltaic power station within the intraday scheduling time period and the historical output power value of the photovoltaic power station are obtained, and the long short-term memory network is used to output the predicted value of the output power increment of the photovoltaic power station within the intraday scheduling time period. The output power prediction value of the photovoltaic power station is determined by the following relationship: ， In the formula, For Always The predicted output power of the photovoltaic power station at any time, is the scheduling time step, For The initial output power value of the photovoltaic power station at the moment, For Always The predicted output power increment of the photovoltaic power station at any time; Taking the minimum difference between the predicted output power of the photovoltaic power station within the daily scheduling time period and the planned output power of the photovoltaic power station in the day-ahead scheduling plan as the optimization goal, an intraday scheduling model is established; The intraday scheduling model satisfies the following relationship: ， In the formula, The difference between the predicted output power of the photovoltaic power station and the planned output power of the photovoltaic power station in the day-ahead scheduling plan is the smallest. For Always The predicted output power of the photovoltaic power station at any time, For the day-ahead scheduling plan The planned output power value of the photovoltaic power station at the time, is an intraday scheduling time period, is the error fluctuation penalty factor of the photovoltaic power station output power; The actual output power value of the photovoltaic power station in the current intraday scheduling time period is obtained as the initial output power value of the photovoltaic power station in the next intraday scheduling time period.

11. The coordinated control method of a distribution network with distributed photovoltaic storage according to claim 1, characterized in that: The MOSEK solver is used to solve the intraday scheduling model, and the optimal solution obtained is the intraday scheduling optimization value within each intraday scheduling time period, which constitutes the intraday scheduling plan.

12. A distribution network coordination control system including distributed photovoltaic storage, characterized in that: include: A data processing module is used to obtain the total load power in the distribution network and the output power of the photovoltaic power station to determine the net load power curve of the distribution network; The net load power curve is divided into different resolution periods according to the gradient and standard deviation, direction change frequency and standard deviation of the net load power in each period during the day-ahead scheduling planning stage. The day-ahead dispatch planning module is used to establish a day-ahead dispatch model with the minimum total cost of distribution network operation as the optimization goal; the fitness of the population is dynamically adjusted in different resolution periods to improve the sparrow optimization algorithm, and the improved sparrow optimization algorithm is used to solve the day-ahead dispatch model. For the optimal solution obtained, the entropy weight method-TOPSIS method is used to determine the day-ahead dispatch plan; The intraday scheduling plan module is used to determine the intraday scheduling time period based on each rolling optimization time scale, taking the maximum value of the time scale that satisfies different resolution time periods at the same time as the rolling optimization time scale; Taking the minimum difference between the predicted output power of the photovoltaic power station within the daily scheduling time period and the planned output power of the photovoltaic power station in the day-ahead scheduling plan as the optimization goal, an intraday scheduling model is established; The optimal solution of the intraday scheduling model is used as the intraday scheduling plan.

Images