A novel coordinated control method and system for multi-type resources in power system

Abstract

The present invention proposes a new type of coordinated control method and system for multiple types of resources in a power system, the method comprising: completing the optimal arrangement of energy storage and the optimal arrangement of photovoltaic unit layout for voltage stability of the new power system; after completing the optimal arrangement, using the approximate characteristics of the Jacobian determinant of the power flow to construct a voltage stability index to capture the voltage fluctuation and safety margin in the system; judging the classification of energy storage working states when the new power system needs emergency control according to the voltage stability index, and determining the output power control under different working states; calculating the load reduction parameters according to the output power control under different states, and adopting the main distribution layered emergency control strategy to realize the coordinated control of multiple types of resources. Based on this method, a new type of coordinated control system for multiple types of resources in a power system is also proposed. The present invention proposes to effectively deal with the volatility problem of the system after large-scale photovoltaic access through the coordinated control of multiple substations, thereby enhancing the stability and recovery capability of the power grid.

Description

Technical Field

[0001] The present invention belongs to the technical field of power system resource regulation and control, and in particular relates to a novel power system multi-type resource coordinated control method and system. Background Art

[0002] With the rapid development of distributed renewable energy, new units represented by photovoltaics and energy storage are connected to the grid on a large scale on the distribution network side, and the traditional power system is clearly transforming to a decentralized new power system. On the one hand, in view of the many environmental and renewable challenges brought by fossil fuels, it is imperative to adjust the energy structure, increase investment in new clean energy such as wind, nuclear and solar energy, and reduce coal-fired power plants in line with the decarbonization and carbon neutrality strategies. On the other hand, further deployment of new energy will have an impact on the stability of the power grid. New energy power generation is characterized by randomness and unpredictable fluctuations. If new units are connected to the grid on a large scale, the randomness of these powers will affect the stable operation of the power system. In addition, the connection of new energy to the grid introduces a large number of power electronic equipment, which in turn will further reduce the system inertia and frequency stability, resulting in a decrease in power quality. Moreover, since the generation, transmission, distribution and consumption of traditional power grids are almost carried out simultaneously, the system capacity must always maintain a dynamic balance. The instantaneous capacity imbalance caused by disturbances will pose a greater threat to the stability, safety and reliability of the entire power grid. Therefore, while the power system is developing in the direction of a high proportion of new energy and decentralization, higher requirements are placed on system resilience and responsiveness, such as point-to-point energy trading, microgrid control, fault detection or demand response.

[0003] In the new power system with multiple types of control resources, the types of units in the system are no longer mainly synchronous machines, but are composed of new energy, energy storage and synchronous machines to supply more flexible new loads. At this time, the operation of the system is still based on the balance of supply and demand. The inherent intermittent nature of new energy units and the uncontrollability of the operation of multiple types of resources have a great impact on the new power system, weakening the development momentum of the new power system and reducing stability and reliability. The new power system needs to integrate different power generation, energy storage and conversion technologies through the coordinated management of different types of resources to achieve the optimized operation of the multi-mode power system and increase the proportion of renewable energy generation.

[0004] To date, low-frequency load shedding is still an important means to maintain frequency stability and avoid frequency drops after system disturbances. However, the existing coordinated control methods for frequency stability of distributed resources focus on the emergency control of a single type of unit or the optimization of traditional machine / load shedding, and cannot reflect the mutual influence of transient characteristics after the aggregation of multiple types of resources, and cannot fully capture the impact of unexpected instantaneous fluctuations on system performance caused by differences in load and renewable energy generation. Therefore, its results cannot fully reflect the actual system performance. Moreover, the simple grading scheme does not take into account the multiple constraints of controlling different types of energy, and often ignores the research on voltage fluctuations that are difficult to avoid after cutting off the photovoltaic units on the distribution network side when shedding loads. Summary of the invention

[0005] In order to solve the above technical problems, the present invention proposes a new method and system for coordinated control of multiple types of resources in a power system, which effectively addresses the volatility problem of the system after large-scale photovoltaic access through coordinated control of multiple substations, thereby enhancing the stability and recovery capability of the power grid.

[0006] To achieve the above object, the present invention adopts the following technical solutions:

[0007] A novel coordinated control method for multiple types of resources in a power system comprises the following steps:

[0008] Establish a power output model for BESS and determine the voltage stability index for transient voltage stability research; use cross entropy to optimize the voltage stability index to determine the top-ranked emergency node set to complete the distribution layout of energy storage units; define the state vector of the combination of photovoltaic power generation prediction and battery storage level in the MDP model, discretize the state vector, determine different types of costs in the MDP model, and calculate the total cost to complete the optimal layout of energy storage and photovoltaic units to enhance system voltage stability;

[0009] After completing the optimal layout of energy storage and photovoltaic units for voltage stability, the voltage stability index is constructed using the approximate characteristics of the Jacobian determinant of the power flow to capture the voltage fluctuation and safety margin in the system;

[0010] Whether the new power system needs emergency control is determined based on the voltage stability index. When control is needed, the energy storage working state is classified according to the shortest discharge time and the remaining discharge time, and the output power control under different working states is determined; the load reduction parameters are calculated based on the output power control under different states, and a main distribution layered emergency control strategy is adopted to realize coordinated control of multiple types of resources.

[0011] A novel multi-type resource coordination control system for a power system, comprising: an optimization arrangement module, a calculation module and a coordination control module;

[0012] The optimization layout module is used to establish a power output model of the BESS and determine the voltage stability index for transient voltage stability research; the cross entropy is used to optimize the voltage stability index to determine the node set of the top emergency situations to complete the distribution layout of the energy storage unit; the state vector of the combination of photovoltaic power generation prediction and battery storage level is defined in the MDP model, and after the state vector is discretized, different types of costs are determined in the MDP model, and the total cost is calculated to complete the layout optimization of energy storage and photovoltaic units to enhance the system voltage stability;

[0013] The calculation module is used to calculate the voltage stability index for capturing the voltage fluctuation and safety margin in the system by using the approximate characteristics of the Jacobian determinant of the power flow after completing the energy storage and photovoltaic unit layout optimization arrangement for voltage stability;

[0014] The coordinated control module is used to determine whether the new power system needs emergency control based on the voltage stability index. When control is needed, the energy storage working state is classified according to the shortest discharge time and the remaining discharge time, and the output power control under different working states is determined; the load reduction parameters are calculated according to the output power control under different states, and the emergency control strategy of the main distribution layer is adopted to realize the coordinated control of multiple types of resources.

[0015] The effects provided in the content of the invention are only the effects of the embodiments, not all the effects of the invention. One of the above technical solutions has the following advantages or beneficial effects:

[0016] The present invention proposes a novel method and system for coordinated control of multiple types of resources in a power system. The method comprises the following steps: establishing a power output model of a BESS, determining a voltage stability index for transient voltage stability research; optimizing the voltage stability index by using cross entropy to determine a node set of emergency situations with a top ranking to complete the distribution layout of energy storage units; defining a state vector of a combination of photovoltaic power generation prediction and battery storage level in an MDP model, discretizing the state vector, determining different types of costs in the MDP model, and calculating the total cost to complete the layout optimization of energy storage and photovoltaic units for enhancing system voltage stability; after completing the layout optimization of energy storage and photovoltaic units for voltage stability, using the approximate characteristics of the Jacobian determinant of the power flow, constructing a voltage stability index for capturing voltage fluctuations and safety margins in the system; judging whether the novel power system needs emergency control according to the voltage stability index, classifying the energy storage working state according to the shortest discharge time and the remaining discharge time when control is needed, and determining the output power control under different working states; calculating the load reduction parameter according to the output power control under different states, and realizing coordinated control of multiple types of resources by using a main distribution layered emergency control strategy. Based on a new type of coordinated control method for multiple types of resources in a power system, a new type of coordinated control system for multiple types of resources in a power system is also proposed. The present invention proposes a multi-resource coordinated control method based on the coordinated layout of energy storage and photovoltaics, voltage stability margin assessment and hierarchical emergency control. The layout optimization, voltage stability characterization and assessment method, and the main and distribution hierarchical emergency control system are three-layer progressive and interdependent system solutions. Layout optimization ensures the rationality of resource distribution and provides a physical basis for voltage stability, thereby achieving real-time monitoring of the system status through characterization and assessment, while the main and distribution hierarchical control quickly dispatches multiple types of resources according to the real-time status when necessary, ultimately ensuring the stable operation of the system.

[0017] By optimizing the layout of energy storage and photovoltaic units, the present invention effectively reduces the impact of photovoltaic fluctuations on the system, significantly enhances the voltage stability and accident resistance of the system, and has better adaptability in scenarios with a high proportion of new energy access. The layout optimization method is superior to traditional passive prevention and control, and can enable energy storage and photovoltaic resources to achieve synergy under accident conditions, thereby improving the grid's resistance to major faults and emergency response level.

[0018] The present invention proposes a voltage stability index based on the Jacobian determinant of the power flow. Compared with the traditional centralized calculation method, it has the advantages of distributed computing, can accurately evaluate the voltage stability margin of the radial distribution network in real time, and significantly reduce the calculation cost. Its high accuracy and scalability ensure the voltage safety performance of the system under active conditions, greatly improving the reliability of the system.

[0019] The main and distribution hierarchical emergency control system designed by the present invention realizes efficient integration and rapid response of multi-regional and multi-type resources through hierarchical coordination of the main control unit and the distribution area unit. The overall strategy can be quickly formulated through the main control unit, while each distribution area unit flexibly coordinates energy storage, photovoltaics and loads, giving priority to the use of energy storage and photovoltaic output to support the stability of the power grid. Compared with the traditional isolated control method, it significantly improves the flexibility and response speed of resource scheduling, providing the power grid with higher recovery capabilities and operational reliability in emergency situations. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] Figure 1 A flow chart of a novel method for coordinated control of multiple types of resources in a power system proposed in Embodiment 1 of the present invention;

[0021] Figure 2 The PQ decoupling control scheme proposed in Example 1 of the present invention;

[0022] Figure 3 A schematic diagram of the discrete and online stages of total cost calculation proposed in Embodiment 1 of the present invention;

[0023] Figure 4 This is a flow chart of coordinated load shedding of multiple zones proposed in Example 1 of the present invention;

[0024] Figure 5 This is a main distribution layer coordination control flow chart proposed in Example 1 of the present invention;

[0025] Figure 6 This is a schematic diagram of a novel power system multi-type resource coordination control system proposed in Example 2 of the present invention. DETAILED DESCRIPTION

[0026] In order to clearly illustrate the technical features of the present solution, the present invention is described in detail below through specific implementation methods and in conjunction with the accompanying drawings. The disclosure below provides many different embodiments or examples for realizing different structures of the present invention. In order to simplify the disclosure of the present invention, the components and settings of specific examples are described below. In addition, the present invention may repeat reference numbers and / or letters in different examples. This repetition is for the purpose of simplification and clarity, and does not itself indicate the relationship between the various embodiments and / or settings discussed. It should be noted that the components illustrated in the accompanying drawings are not necessarily drawn to scale. The present invention omits the description of known components and processing techniques and processes to avoid unnecessary limitations on the present invention.

[0027] Example 1

[0028] Embodiment 1 of the present invention proposes a novel method for coordinated control of multiple types of resources in a power system, aiming to solve the complexity of coordinated control of multiple types of resources in an active distribution network scenario. Figure 1A flow chart of a novel method for coordinated control of multiple types of resources in a power system proposed in Embodiment 1 of the present invention;

[0029] In step S1, a power output model of the BESS is established to determine the voltage stability index for transient voltage stability research; the cross entropy is used to optimize the voltage stability index to determine the node set of the top emergency situations to complete the distribution layout of the energy storage unit; the state vector of the combination of photovoltaic power generation prediction and battery storage level is defined in the MDP model, and after the state vector is discretized, different types of costs are determined in the MDP model, and the total cost is calculated to complete the optimal layout of energy storage and photovoltaic units to enhance system voltage stability.

[0030] First, the power output model of the battery energy storage system (BESS) is used to study the transient voltage stability. Since the switching speed of power electronic devices is generally much faster than the electromechanical response of synchronous generators, the BESS is modeled using a first-order transfer function. Figure 2 This is the PQ decoupling control solution proposed in Example 1 of the present invention.

[0031] The power output model of BESS is established as:

[0032] ; (1)

[0033] ; (2)

[0034] in, is the power output of BESS; is the time constant of the BESS power converter, which is generally much smaller than the inertia time constant of large generators and is set to 0.02s; represents the reference power; represents the time step; Represents the current state of charge of the BESS; Represents the maximum value of the BESS state of charge; Represents the minimum value of the BESS state of charge; Represents the maximum limit of the power output of the BESS; is the controller gain of each BESS, which represents the theoretical maximum active power output of the BESS during transients and is set to 10.0; Represents the system frequency change.

[0035] Design the voltage stability index (VSI) for transient voltage stability research, assuming that there is a previous Serious failures need to be handled, for node indicators ,exist The percentage change of voltage at time is:

[0036] ; (3)

[0037] in, For the The percentage of voltage change after a fault occurs; Node index exist Failure Voltage at the moment; Node index Initial voltage when no fault occurs; is the total number of node indicators; is the node indicator; is the maximum number of time steps;

[0038] The severity indicators are defined as:

[0039] ; (4)

[0040] in, For the The severity index of each fault; For Node exist Fault time Severity index at the time of is the total number of nodes;

[0041] The bigger, the The next step is to establish a measure to quantify the contribution of BESS to voltage improvement after a fault, using the maximum voltage recovery sensitivity for voltage stability index calculation:

[0042] ; (5)

[0043] in, is the voltage stability index, i.e., the sensitivity of the voltage recovery degree to the theoretical maximum power output of all placed BESS units; For Node exist The voltage value before the fault; For Node exist The voltage value after a fault; For all BESS power converters ;

[0044] No. Failure No. The overall voltage stability index at each node is:

[0045] ; (6)

[0046] The overall voltage stability index for all fault configuration combinations is:

[0047] ; (7)

[0048] Assume that there are a total of BESS units are placed to optimize the voltage stability index, then the optimal placement problem of multiple BESS units is:

[0049] ; (8)

[0050] ; (9)

[0051] in, is a binary variable whose value “1” means that a BESS will be placed at the node superior; The number of BESS units to be placed to achieve the optimal voltage stability index; Represents the voltage stability margin.

[0052] The process of optimizing the voltage stability index by cross entropy to determine the top-ranked node set of emergency situations and completing the distribution layout of energy storage units includes: extracting samples based on probability distribution; and updating the parameters of the probability distribution of the extracted samples based on the extracted samples.

[0053] Similar to the CE method for rare event sampling, the estimation-based CE method is applied to the optimization problem, assuming that the optimal solution obeys a potential probability distribution, which can be expressed by a function Then, the process of finding the optimal solution is simulated by continuously sampling the "rare events", that is, finding the event of the theoretical global optimal solution. Specifically, for the following maximization problem:

[0054] ; (10)

[0055] in, is the objective function; is a decision variable that appears in the form of a random variable

[0056] Greater than Probability for:

[0057] ; (11)

[0058] in, for Greater than The probability of (hereinafter referred to as or ) is an "indicator function". If the event in the brackets occurs, then The value is "1". Adopting the idea of ​​"importance sampling", assuming is another probability function, and when hour, ,but:

[0059] ; (12)

[0060] It is deduced that:

[0061] ; (13)

[0062] in, for The weighted average of .

[0063] The next step is to find a value as close as possible to of You can choose To approximate ,remain To solve it. To this end, the cross entropy minimization based on "KL divergence" is adopted:

[0064] ; (14)

[0065] ; (15)

[0066] in, is the new reference parameter for the second round of importance sampling; and Indicates that the error is within the acceptable range and The approximate value of represent and The KL divergence between them represents the error between them. When it is an exponential family distribution similar to Gaussian distribution, we can usually get For the optimization problem, a multi-layer CE method is used, that is, in each iteration, , For sequence , ..., ( is the number of samples) Quantile, that is:

[0067] ; (16)

[0068] in, is a predefined ratio called the “Elite Retention Rate”. of Using formula (17) on a random sample Updated to Iterate until Convergence or maximum number of iterations reached.

[0069] ; (17)

[0070] in, The status before the update; is the state after iteration; It is decision variables; is the number of samples.

[0071] For combinatorial optimization problems, the steps to apply the CE method are as follows:

[0072] Assume that the density function The parameters are The multivariate Bernoulli distribution of is:

[0073] ; (18)

[0074] in, is the density function; is a multidimensional state variable, ,Depend on composition.

[0075] is a multidimensional state variable, ,Depend on composition, represent Probability of occurrence.

[0076] From formula (17), we can know that: Samples :

[0077] ; (19)

[0078] in, for The weighted average of .

[0079] At this point, the distribution layout of the energy storage unit is completed. In practical applications, each time a potential combination of BESS locations is generated in the CE method, the VSI calculation is first performed, and then the set of nodes with emergency situations with the highest SI values ​​in the layout results is comprehensively considered to obtain the selected location and capacity. On this basis, the MDP-based photovoltaic and energy storage layout strategy is proposed next.

[0080] Construct an MDP model and define the relative error of photovoltaic power generation prediction in the MDP model ( ) and the battery storage level of each BESS ( ) process is:

[0081] ; (20)

[0082] in, is the relative error of the PV power generation prediction of the first DER; For the The relative error of the PV power generation forecast for each DER; is the battery storage level of the first BESS; For the The battery storage level of each BESS; is the number of DERs; is the number of BESS units; the state space is discretized by dividing the value into intervals. The relative error of the PV power generation prediction of each DER is independent of the action and is considered random.

[0083] To limit the values ​​to the range of -1 to 1, the relative error of the PV power generation forecast for each DER is defined as:

[0084] ;(twenty one)

[0085] in, is the current output level of the photovoltaic unit; is the reference power output of the unit; is the relative error of the PV power generation prediction for each DER;

[0086] The battery storage level is expressed as a relative value in the range of 0~100%, which is assumed to be a constant value. It is assumed that the distribution system operator (DSO) determines the amount of photovoltaic power generation reduction and reactive power injection for each DER, the charging and discharging of the DER battery of each BESS, and the load shedding. The system action can be expressed as:

[0087] ;(twenty two)

[0088] in, is the action vector; is the reduction in PV power generation of the first DER; For the PV generation reduction per DER; The injected reactive power of the first DER; For the The injected reactive power of each DER; Charge and discharge the DER battery of the first BESS; For the The DER batteries of each BESS are charged and discharged; is the load shedding amount;

[0089] For each load value, there is an assigned number of nodes that are affected by the load shedding. Other types of actions are expressed as relative values ​​and divided into equal intervals.

[0090] Assuming that the battery energy storage state transition is deterministic, therefore, =0 or 1, that is, under the action of BESS, the BESS state Transfer to state For the nighttime period, the PV power generation is equal to zero, so the model is completely deterministic during this period. The stochastic method is only applicable to PV power generation during the daytime period. to another The transition probability of a state is calculated as the ratio of the number of times that transition occurs to the total number of times that particular state occurs:

[0091] ;(twenty three)

[0092] in, For photovoltaic units from the state arrive The probability of For the first state; For the second state; is the number of times the transfer occurred; for The total number of times the state occurs;

[0093] Each unit and BESS The transition probabilities are independent of each other, so the obtained transition probability is equal to the product of all probabilities; accordingly, from the third state To the fourth state In action The transition probability under is as follows:

[0094] ;(twenty four)

[0095] in Represents the energy storage unit from the original state Transfer to state The probability of Represents the photovoltaic unit from the original state Transfer to state The probability of By Status To status In action The transition probability under .

[0096] The process of determining different types of costs in the MDP model and calculating the total cost to complete the optimal layout of energy storage and photovoltaic units to enhance system voltage stability includes:

[0097] The total cost in time is:

[0098] ; (25)

[0099] in, represent Total cost of time; represents the equivalent cost of underutilization of PV resources; Represents the value of abandoned electricity; represents a penalty for load shedding; Represents the load shedding value; Represents the load shedding loss on the power grid side; Represents the load loss on the side of distributed energy resources; Represents the equivalent network loss cost of the power grid; represents the equivalent power loss cost of distributed energy resources; for Flexibility of services provided; For active power transfer between distribution and transmission systems; For reactive power transfer between distribution and transmission systems;

[0100] Figure 3 Schematic diagram of the discrete and online stages of the total cost calculation proposed in Example 1 of the present invention; the offline calculation is performed one day in advance. The result of these calculations is the expected utility value for each epoch and state. Assume that the following information is known one day in advance: the load demand forecast for each node, the photovoltaic power generation forecast for each DER, and Plan: Power delivery constraints and flexibility service requirements, as well as appropriate information on the rewards for providing the services. For each time period, state, and measure, a cost value is calculated and the constraints are checked. If a constraint is violated, the cost value is replaced by a very large number. The expected utility is then calculated for each time epoch and state using backward induction:

[0101] ; (26)

[0102] in, Calculate cost values ​​for each time period, state, and measure; Represents offline computing Expected utility at a given moment; represents the penalty coefficient, which represents the penalty cost after violating the constraint conditions; represent At any time, the state To status In action The transition probability under ; Represents the expected utility in the next period; Represents the possible state set of offline computing presets.

[0103] Online calculation: Due to inaccurate load demand forecast, The provided planning flexibility values ​​and the quantified values ​​of power generation are inaccurate, and offline calculations will inevitably produce errors. In order to reduce the impact of these inaccurate estimates, the instantaneous cost of each action is recalculated using actual data. Assume that at each time epoch in real time, Each will provide information on load demand, actual value of PV generation, transmission system constraints and flexibility service needs. All actions other than have no effect on the state transition probability, so the action range of the current epoch can be extended without modification ( )

[0104] ; (27)

[0105] in, Represents the instantaneous cost during online computing; Represents the actual set of actions that may occur during online computing; represents the immediate cost for each action; represent At any time, the state To status In action The transition probability under ;

[0106] At this point, the optimized layout of energy storage and photovoltaic units to enhance system voltage stability has been completed.

[0107] The layout optimization method in the present invention lays the foundation for the stability of the power grid. In the power system, the volatility of photovoltaic power generation and the response speed of the energy storage system are inconsistent, which may lead to fluctuations in system frequency or voltage. Therefore, the layout optimization method comprehensively considers the power generation characteristics and response speed of energy storage and photovoltaic units, and reduces the impact of photovoltaic fluctuations on the stability of the power grid by reasonably selecting the location and capacity setting of the units. This layout design not only enhances the anti-interference ability of the power grid under fault conditions, but also provides stable resource configuration and response conditions for subsequent voltage stability characterization and emergency control. At this stage, the distributed layout of photovoltaics and energy storage determines the availability of control resources and their response efficiency in subsequent steps. In other words, reasonable layout optimization ensures that when an accident occurs, various types of resources can effectively participate in subsequent stability control and response, laying the foundation for the overall stability and reliability of the system.

[0108] In step S2, after completing the optimal layout of energy storage and photovoltaic units for voltage stability, the voltage stability index is constructed using the approximate characteristics of the Jacobian determinant of the power flow to capture the voltage fluctuation and safety margin in the system;

[0109] The process of constructing a voltage stability index to capture voltage fluctuations and safety margins in the system using the approximate characteristics of the power flow Jacobian determinant includes:

[0110] First, establish the distribution network model. is a directed tree representing a radial distribution network, where Each node in represents an actual node. Each edge in represents a route. Note that . A directed edge in ,express yes For each node ,make represents the set of all its child nodes. Node 0 represents the root of the tree, corresponding to the distribution network substation. For each ,make Its only parent.

[0111] Now define the basic variables. For each ,make For Node To Node The square of the complex current amplitude, For Node To Node The sending end complex power is For Line The complex impedance on the node. ,make For Node The square of the complex voltage amplitude at For Node The net complex power demand (load minus generation) at . Finally, we use the notation 1 for the vector of all 1s and 0 for the vector of all 0s. Finally, we use the notation 1 for the vector of all 1s and 0 for the vector of all 0s.

[0112] Modeling distribution networks using relaxed branch power flow equations:

[0113] ; (28)

[0114] in, Represents the active power of the node; Representative Node The total active power to the adjacent nodes; Representative Node resistance to adjacent nodes; Representative Node The square of the complex current amplitude to the adjacent node; let Represents the set of all its child nodes; Representative branch Active power; Represents the node reactive power; Representative Node The sum of reactive power to adjacent nodes; Representative Node reactance to adjacent nodes; Representative branch Reactive power; Representative Node The square of the complex voltage amplitude at ; Representative Node The square of the complex voltage amplitude at ; Representative branch resistance; Representative branch Active power; Representative branch Reactance; Representative branch Reactive power; Representative Node To Node The square of the branch complex current amplitude;

[0115] The relaxed vector form of the branch flow equation is:

[0116] ; (29)

[0117] in, , and , is the defined phasor, respectively, by the scalar , and , Stacking results; and Represents the two (0, 1) matrices defined in this paper, where For each line Select branch The matrix of For each line Select branch Matrix of represent Eliminate The matrix obtained by row; represent The transposed matrix of Represents an n×n matrix, with and The phasors obtained by multiplication are diagonal elements and zero elsewhere; Represents the resistance matrix, which is an n×n matrix, and the resistance on the diagonal and zero elsewhere; Represents the reactance matrix, which is an n×n matrix. on the diagonal and zero elsewhere; Represents the branch active power matrix, which is an n×n matrix. on the diagonal and zero elsewhere; Represents the branch reactive power matrix, which is an n×n matrix. on the diagonal and zero elsewhere; represents the square of the complex current amplitude; represents the square of the complex voltage magnitude; node 0 is modeled as a relaxed bus, where All other nodes are modeled as PQ buses, where complex power demands (real and reactive) are imposed and are independent of bus voltage. Interpreted as system parameters, the relaxed branch power flow model is Variables specified equation, that is .

[0118] Next, we study the voltage stability characterization and propose stability indices. This section first proposes an approximation method for simplifying the Jacobian determinant of the power flow, which can be used for scalable and distributed computing when the measured data of the grid variables are available. Then, based on this approximation, we propose a voltage stability index to quantify the distance of the power system from voltage collapse.

[0119] The load limit of the power system is the critical operating point of the grid (determined by the node injection power), at which the power transfer reaches the maximum value. After this point, the grid is overloaded and the relaxed branch power flow equation has no solution. The system has multiple load capacity limits, corresponding to different demand configurations. Ideally, the power system will operate away from these points with sufficient safety margin. On the other hand, the voltage stability point of the power system, and ; The power system usually operates near this point. The Jacobian matrix of the power flow equation is formalized below to represent the standard reasoning of the load capacity limit, and these results are used for the branch power flow model adopted. Consider two vectors:

[0120] ; (30)

[0121] in, and Represents two vectors, corresponding to system variables and system parameters respectively; and Represents the active and reactive power flows of each branch; Represents the square of the branch complex current amplitude; represents the square of the node complex voltage amplitude; and Represents the active and reactive power of the node load; Represents the system voltage; Represents the system active power flow; Represents the reactive power flow of the system; the reactive power flow is the trend of reactive current flowing from the power source into the loads of each component of the system; represents the first spatial set; represents the second spatial set;

[0122] The two correspond to system variables and system parameters respectively. Formula (29) is expressed in implicit form as follows:

[0123] ; (31)

[0124] The loadability limit corresponds to the scalar function The maximum value of the physical grid constraint is the set

[0125] ; (32)

[0126] Directly apply the KKT optimal condition to obtain the flow Jacobian Singular under load limit, i.e. Based on this, the standard representation of grid voltage stability is adopted, which is given in the following definition: A network described by a relaxed branch power flow model with 1 relaxed node and The voltage stability domain of the distribution network with a PQ node is an open area where the power flow solution set satisfies:

[0127] ; (33)

[0128] Although there may be other feasible regions where the determinant of the power flow Jacobian is negative, the region characterized by (33) corresponds to the actual operating point of the power system operation. When the branch flow model is adopted, Take the form:

[0129] ; (34)

[0130] in, and They are respectively and The matrix obtained by removing the first row is and They are and The transposed matrix of is the first canonical basis vector; Represents a The all-zero area means that all elements in this area are 0;

[0131] Formula (34) is simplified to the power flow Jacobian matrix:

[0132] ; (35)

[0133] in, represents the reduced-order power flow Jacobian; It is an effective tool for characterizing the voltage stability domain and analyzing the voltage stability of distribution networks. When studying singularities, studying the simplified power flow Jacobian is completely equivalent to studying the original power flow Jacobian and provides a more accurate characterization of the power grid voltage stability region. If the root (node ​​0) has multiple child nodes, the dimension of the problem of calculating the power flow Jacobian determinant can be further reduced.

[0134] Reduced Order Power Flow Jacobian Direct inspection of shows that for realistic parameter values ​​and operating conditions, its off-diagonal elements (especially its lower diagonal elements) are significantly smaller than its diagonal elements. The voltage stability region is defined as In practical applications, grid operators need to determine a threshold , and apply As a practical voltage stabilization measure. In order to fully utilize the capacity of the power grid, it is necessary to select The value of When , the operation point of the power grid is very close to a load capacity limit. From the numerical experiments, it can be seen that The reasonable choice of depends essentially on the size of the network. In the voltage solution, All eigenvalues ​​of are equal to 1. As the power demand increases, the eigenvalues ​​begin to move toward the origin. Since the number of eigenvalues ​​is equal to The size of is equal to the size of the network. Obviously, larger networks are associated with exponentially smaller determinants. Based on this, the following As an indicator of voltage stability;

[0135] ; (36)

[0136] in, It is the actual voltage stability index;

[0137] For a certain threshold , the actual voltage stability measure becomes:

[0138] ; (37)

[0139] in, is the minimum value of the voltage stability index;

[0140] The voltage stability index is approximately defined as:

[0141] ; (38)

[0142] in, It is the approximate value of voltage stability index; for The product of the elements on the diagonal;

[0143] The approximate interval of the voltage stability index is:

[0144] ; (39).

[0145] The voltage stability characterization and evaluation method in the present invention is a real-time monitoring and judgment method based on layout optimization. Different from general voltage monitoring, this method combines the characteristics of active distribution network, utilizes the approximate characteristics of the Jacobian determinant of the power flow, and constructs a voltage stability index with high scalability and distributed computing advantages. Since the layout optimization link has reasonably arranged the positions of photovoltaic and energy storage units, this characterization and evaluation method can more accurately capture the voltage fluctuations and safety margins in the system, and thus provide the system with more accurate stability evaluation results. This indicator not only evaluates the voltage stability of each area in real time, but also can issue early warnings for the trend of voltage instability when encountering sudden faults, and provide real-time data support for the main distribution layered emergency control of the system. As a result, a closed loop is formed between layout optimization and stability evaluation: the optimized resource allocation directly improves the accuracy of the characterization, and the real-time characterization and evaluation feed back the status information to the emergency control module to prepare data for subsequent responses.

[0146] In step S3, it is determined whether the new power system needs emergency control according to the voltage stability index. When control is needed, the energy storage working state is classified according to the shortest discharge time and the remaining discharge time, and the output power control under different working states is determined; the load reduction parameters are calculated according to the output power control under different states, and the emergency control strategy of the main distribution layer is adopted to realize the coordinated control of multiple types of resources.

[0147] First, the control of energy storage units is analyzed. In order to achieve refined control of energy storage units, the different charge states of each energy storage are comprehensively considered to determine whether it can participate in emergency control, so as to avoid the situation where insufficient power affects the energy storage power support. First, the definition of the shortest discharge time and the remaining discharge time is proposed to measure the energy storage state, and the energy storage is divided into four working states.

[0148] Remaining discharge time Longer than the shortest discharge time When the energy storage capacity is sufficient, it is the first working state; the remaining discharge time The duration of energy storage at maximum power output; the shortest discharge time Refers to the shortest time that the grid requires energy storage to provide power support in an emergency;

[0149] Remaining discharge time Shorter than the shortest discharge time When the energy storage capacity is insufficient, if it is in the charging state, it is the second working state of the energy storage;

[0150] In standby or discharge state, if the battery is insufficient but can provide power support, it is the third working state;

[0151] If the battery level is lower than the level that makes it difficult to support power, the fourth working state is entered, and the control amount is added to stop charging and discharging.

[0152] For the energy storage in the first state, the power shortage is supported by discharging at the maximum output within the system required time;

[0153] ; (40)

[0154] in, Energy storage for the first working state Output power when emergency power support is needed; for The remaining energy storage capacity at any time; The minimum allowable remaining power to provide emergency support; is the energy storage discharge efficiency;

[0155] The second working state in the charging state is converted into the third working state or the fourth working state according to the charging state and the delay requirement of the unit;

[0156] The energy storage in the third state has insufficient remaining discharge time to support the shortest discharge time required by the system at the maximum discharge power. If the maximum power output command is forcibly applied at this time, it is very likely that the power will drop due to insufficient energy before the shortest discharge time, resulting in a new power shortage. The system frequency, which is already facing a greater threat, will further deteriorate. Therefore, the power output is reduced to achieve the shortest discharge time.

[0157] In the third working state, the power output is reduced to achieve the shortest discharge time, and the output power control is:

[0158] ; (41)

[0159] in, Energy storage for the third working state Output power when emergency power support is needed; The charge that can be maintained for the shortest power supply time under the maximum discharge state; is the maximum amount of energy stored;

[0160] For the energy storage in the fourth state, the charge is too low to provide power support, so control is applied to stop charging and discharging.

[0161] The process of calculating the load reduction parameters according to the output power control in different states includes: calculating the frequency change rate of the system; calculating the minimum peak occurrence time after the system has a power shortage according to the frequency change rate of the system; calculating the maximum frequency deviation, the minimum frequency and the stabilized frequency according to the minimum peak occurrence time; through the improved system frequency response model, combined with the initial frequency change rate of each node monitored by the wide area measurement system, the actual power shortage caused by the system failure is calculated, and then the load shedding amount of the unit's primary frequency regulation is calculated.

[0162] Next, we combine the transfer function model mentioned above, comprehensively consider the load frequency regulation effect, load characteristic response curve and unit frequency regulation characteristics, and calculate and analyze the various parameters of load reduction. First, consider the calculation of power shortage. When the system is disturbed, the initial value of the frequency change rate, that is, time, system (Rate of Change ofFrequency) is:

[0163] ; (42)

[0164] in, is the frequency change rate of the system, that is, how fast it changes; is the voltage change; is the power change; is the voltage response coefficient; is the system inertia.

[0165] Reverse solving for the power deficit yields:

[0166] ; (43)

[0167] The above formula can be used to calculate the actual power shortage based on the measured frequency change. It can be seen that voltage fluctuations will instantly affect frequency changes, and the power shortage calculation is affected by the coupling effect of unit inertia and load characteristics. If the system constant power load accounts for a small proportion, ignoring the voltage characteristics will mistakenly believe that the frequency drop is not suppressed, thereby underestimating the actual power shortage. In addition, the initial It will not be affected by the load frequency regulation effect and the primary frequency regulation parameters of the unit. This is because at the moment when the power shortage occurs, the system frequency remains unchanged due to inertia.

[0168] When the system has a power shortage, the system frequency gradually decreases and the primary frequency modulation starts. Here, it is assumed that the system has sufficient spare capacity and the low-frequency load reduction is not considered. The time when the lowest frequency peak occurs is calculated. According to the frequency change curve, when the frequency is at the lowest point, the frequency change rate is 0, so:

[0169] ; (44)

[0170] From the above formula, we know that the time when the lowest point occurs has nothing to do with the power shortage caused by the fault. Therefore, if the node voltage can remain unchanged, it will have no effect on the lowest point time. Substituting the lowest point time into the formula, we can get the maximum frequency deviation:

[0171] ; (45)

[0172] From the above formula, we can know that the regulation effect coefficient, damping coefficient and load regulation effect coefficient of primary frequency regulation all play a supporting role in frequency drop. Low-frequency load shedding needs to judge the severity of the fault through the maximum frequency deviation. If the load is ignored, If the parameters are not set correctly, the severity of the accident will be overestimated. In some cases, the load reduction device will malfunction, resulting in unnecessary losses.

[0173] According to the final value theorem, the frequency steady-state offset value is:

[0174] ; (46)

[0175] in, is the maximum frequency deviation;

[0176] Start-up criteria for coordinated control. According to the technical regulations for automatic low-frequency load shedding in power systems, the start-up value for low-frequency load shedding in power systems should be high. This is because low-frequency load shedding is the last line of defense to maintain the safety and stability of the power grid. When there is a serious power shortage and no capacity available, it is necessary to disconnect appropriate loads to prevent the frequency from continuing to drop. A higher starting frequency allows the load shedding device to disconnect a certain load at the beginning of the frequency drop, effectively raising the lowest point of the system frequency and reducing the risk of system instability.

[0177] However, in the case of small-scale power shortage faults, the load regulation effect and the primary frequency regulation characteristics of the unit should be fully utilized, as well as some methods that do not cause load shutdown to adjust the frequency. And the rapid regulation of new energy sources for low-frequency pump cutting in energy storage power stations, etc. If the starting frequency is set too high, the system backup and other adjustment measures will not be fully utilized, causing the load shedding device to malfunction in some situations where load shedding is not required, resulting in unnecessary load shedding losses.

[0178] Taking into account the low-frequency tolerance of the units and loads, the safety of the system and the speed of the spinning reserve start-up, the low-frequency load shedding start-up round in my country is generally set at 49.0Hz. And under the premise of not considering low-frequency load shedding, the power shortage value when the lowest point of the transient frequency of the system is higher than 49.0Hz and the steady-state frequency is higher than 49.5Hz is defined as the system reference power shortage.

[0179] According to the load frequency response model mentioned above, at the beginning of the frequency drop, the frequency evolution of the inertia center of the power system is calculated based on the wide-area measurement system. The lowest point and the stable frequency after the accident are calculated as follows:

[0180] ; (47)

[0181] ; (48)

[0182] in, is the lowest frequency, and the frequency after stabilization; In order to combine the power response model of the load, the linear relationship parameters of power and voltage change are obtained by taking the partial derivative of the voltage near the rated point; It indicates the ratio of the system frequency modulation active increment to the frequency deviation. The larger it is, the smoother the power-frequency characteristic of the unit is, and the frequency is closer to the rated level after the balance is restored. It indicates the damping coefficient of the unit. When the system speed decreases, the loss caused by damping decreases, which can alleviate the power shortage to a certain extent. It is the coefficient of power compensation and frequency deviation in load frequency characteristics. This is the adjustment time constant caused by valve opening, etc. Based on the power shortage, it is estimated from the frequency drop speed and system inertia. , , , are system parameters, which are:

[0183] ; (49)

[0184] Based on the frequency of each node measured by the wide area measurement system and the control center Model, if the predicted system frequency minimum point is lower than 49.0Hz (unit value is less than 0.98), or the frequency steady-state value does not reach 49.5Hz (unit value is less than 0.99) after frequency recovery, it is judged that the power shortage has reached the load reduction starting value.

[0185] Therefore, the load shedding start criterion of the adaptive low-frequency load shedding strategy considering load characteristics is:

[0186] ; (50)

[0187] From the above analysis, we can see that through the improved system frequency response model, combined with the initial frequency of each node monitored by the wide area measurement system , calculate the actual power shortage caused by system failure.

[0188] The system expected recovery frequency Set to 49.8Hz. Under the condition of ensuring the normal operation of the system, make full use of the frequency regulation effect of energy storage output and load to reduce the load shedding. In addition, the load shedding amount should also consider the spare capacity of the system and the use of other regulation measures. Considering the load regulation effect and the load shedding amount of the unit's primary frequency regulation, the calculation is as follows:

[0189] ; (51)

[0190] Solving for the load shedding amount yields:

[0191] ; (52)

[0192] in, Indicates the load shedding amount, Indicates the magnitude of the power shortage disturbance; Indicates The rated active power of the node load is represents the number of load nodes, Indicates the active power released by other regulation measures.

[0193] According to the dynamic principle of power system frequency, any change in unit output will cause a change in system frequency. When the unit output increases or the low-frequency load reduction device is activated, the system power shortage is reduced accordingly, resulting in a step-like increase in the system frequency change rate. Conversely, a decrease in unit output and a sudden increase in load will cause the system frequency change rate to decrease, thereby accelerating the frequency drop.

[0194] The coordinated load shedding of multiple stations calculates the actual power shortage through the initial frequency change rate and removes it at one time. If the load shedding is appropriate, the frequency will rise until it finally stabilizes at the expected recovery frequency. However, in actual situations, there are often various operating conditions and uncertainties in predictions, resulting in errors between the calculated power shortage and the actual power shortage. This may lead to two situations that are not in line with expectations: 1. After the round of load shedding, the frequency continues to drop. This indicates that a shortage has occurred, that is, the power shortage calculation is too small. At this time, the electromagnetic power of the system is still less than the mechanical power, and the rotor is in a deceleration state. At this time, it is necessary to estimate the system power shortage again based on the frequency change rate after load shedding, and perform the next round of shedding. 2. After the round of load shedding, the frequency begins to recover, but it does not recover to the expected recovery frequency in the end, and it hovers below 49.5Hz and above 49.0Hz for a long time. At this time, special wheel action needs to be considered. The special wheel determines the load shedding amount through offline setting, and by removing a small amount of load, the frequency hovering below 49.5Hz is restored to the expected recovery frequency.

[0195] When shedding the load, it is necessary to consider the frequency and voltage response characteristics of the load. Shedding the appropriate load can speed up the recovery of the system frequency, or reduce the total amount of load shedding while ensuring the recovery speed. From the analysis of the load frequency response characteristics in the previous section, it can be seen that the frequency regulation effects of loads in different regions are different. Among them, loads with large frequency regulation effects will reduce active power consumption when the frequency drops, and have a suppressive effect on the frequency drop. Therefore, loads with smaller load regulation effects should be given priority when they are shelved. This increases the equivalent regulation effect of the remaining loads and has a stronger ability to suppress frequency drops.

[0196] Under-frequency load shedding needs to consider the importance of the load, so as to determine the priority of load shedding and load removal. For example, hospitals, large chemical enterprises, etc. will have serious consequences if the power is cut off. These loads should not be removed when the power shortage occurs. The loads of the power system are divided into three levels according to their importance. The first-level load is the load that will threaten life safety and cause major losses in politics and economy once the power supply is interrupted. The second-level load is the power load that will cause equipment damage, product scrapping and take a long time to recover once the power is cut off. Loads that do not belong to the above two categories are classified as the third category. Under-frequency load shedding should give priority to cutting off the third category load and ensure stable power supply to the first category load. When choosing to cut off the load, users who can be temporarily cut off also need to be coordinated by the control center. According to the specific situation of the load, the load shedding should be arranged in an orderly and planned manner to minimize the economic losses caused by cutting off the user load.

[0197] On this basis, the selection of load shedding locations needs to comprehensively consider the load importance and the load changes caused by the frequency regulation effect. Priority is given to the shedding loads with low frequency response characteristics among the three-level loads, and the shedding loads are gradually removed from low to high according to the importance and load frequency response characteristics. Based on the derived load shedding parameters, a load shedding scheme that considers the coupling characteristics of the energy storage output and load frequency on the distribution network side is proposed. The flow chart is as follows Figure 4 As shown, Figure 4The flowchart of multi-area coordinated load shedding proposed in Example 1 of the present invention is as follows: 1. The bus voltage and the frequency change rate of each node are collected through the wide-area measurement system. 2. The power of each load and unit is converted, and the actual power shortage and the maximum frequency deviation are calculated according to the above formula. 3. Monitor the current frequency of the system. If it is found that the trigger condition has been reached, it means that the low-frequency load shedding does not need to be performed, and go to step 1 to continue monitoring the system frequency. Otherwise, go to step 4. 4. The energy storage units in each area quickly increase their output. At the same time, the load regulation effect and the system spare capacity are considered to calculate the load shedding amount. If the frequency is restored, go to step 3, otherwise go to step 5. 5. According to the load priority and frequency-voltage characteristics, the load shedding amount is allocated to the appropriate area for load shedding. On this basis, the main distribution layered coordinated control of the system is proposed. With the access of new power systems to a high proportion of distributed photovoltaics and important loads of different levels, the active characteristics of the distribution network are becoming more and more obvious. The traditional emergency load shedding strategy has poor adaptability to the increasingly closely interconnected power grid, resulting in the disconnection of distributed photovoltaics or important loads, which then causes more serious chain failures. In this context, combined with the multi-station coordinated control strategy proposed in the previous section, a main distribution layered coordinated control load shedding strategy is proposed. The operating parameters of the main distribution network are analyzed and extracted, and a load shedding strategy based on energy storage output and load response characteristics is proposed at the main network layer. The load shedding amount of each distribution network in an emergency scenario is obtained through the main network load shedding optimization model. The distribution network substation receives the instruction, and constructs the optimal load shedding model of the distribution network by analyzing the importance of each station and the minimum severity of the load shedding consequences. The multi-station load shedding coordinated control model proposed in the previous section is applied to remove the load step by step from low to high according to the load importance and load frequency response characteristics. The main distribution layered coordinated control process is as follows: Figure 5 shown. Figure 5 This is the main distribution layered coordinated control flow chart proposed in Example 1 of the present invention; a coordinated load shedding strategy is proposed at the main distribution layer, and the dual triggering of event triggering and frequency response is completed by constructing a main grid dual-target load shedding model. When the power grid encounters a serious fault and load shedding measures need to be taken, the cost of restoring system stability should be as small as possible to avoid excessive accident penalties. A load shedding optimization model that takes into account the load frequency response characteristics and is supported by the energy storage output on the distribution network side is established to reduce the risk of the power grid as much as possible while ensuring that the load shedding action of the power grid is minimized, and under this goal, network flow constraints, voltage constraints, load shedding constraints, line safety constraints, thermal power unit output constraints and other conditions are met, taking into account both safety and economy. The specific process is as follows:

[0198] In an emergency, the main control unit will solve the load shedding amount of each distribution network The command is sent to the distribution network substation, which quickly calculates the three-level interruptible load capacity that can be used to reduce the power command by adjusting the gear position. Secondary interruptible load capacity that can only be achieved by controlling the switch to reduce power command , the load frequency response characteristics are used to gradually cut off the load from low to high. If the interruptible load capacity still does not meet the load reduction requirements, the load cutting amount required for the subsequent load cutting action is pass Calculate, where In order to reduce the interruptible load, the optimal load shedding model of the distribution network is used to solve the load shedding location. At the main grid level, a dual-driven load shedding strategy based on events and responses is proposed, which solves the problem that the traditional load shedding strategy is not adaptable enough to unknown serious faults encountered by the power grid, and effectively avoids more serious chain failures caused by inaccurate calculation of load shedding or mis-cutting when facing unknown emergency faults; at the distribution network level, the frequency response characteristics of the load are considered for load shedding design, and the optimal load shedding model of the distribution network is constructed from low to high by analyzing the importance of feeders and the frequency response characteristics of the load. When selecting the load shedding location, the impact of load shedding on the power grid is taken into account, which reduces the impact of distributed photovoltaic or important load disconnection on the power grid while reducing the risk of the power grid, and improves the safety and stability of the power grid.

[0199] The dual-drive load shedding strategy based on events and responses proposed by the main control unit summarizes and analyzes the monitored frequency information, and estimates the frequency deficit in combination with system information after the trigger conditions are met. The load shedding amount is allocated to the response drive module according to the load importance and frequency response characteristics of multiple stations in each feeder. The objective function of the load shedding optimization model at the main grid level is:

[0200] Grid load shedding Minimum:

[0201] ; (53)

[0202] in, is the total load, For Node The load capacity, For Node Required load shedding ratio.

[0203] Comprehensive risk of power grid Minimum:

[0204] ; (54)

[0205] in, Balancing risks for the power grid, For power grid security risks, Absorb risks for new energy consumption; is the weight coefficient corresponding to the grid balance risk; is the weight coefficient corresponding to the power grid security risk; is the weight coefficient corresponding to the risk of new energy consumption.

[0206] The master layer After being transmitted to the substation, the substation first calculates the , and interruptibility factor ,when When the frequency response characteristic is low, only the downshift command needs to be issued to the load part with lower frequency response characteristic to execute the load shedding action; if All the third-level interruptible loads will be cut off, and the shortage of load cut will be made up according to the second-level interruptible loads with higher priority in the load response characteristics; if , then all current interruptible loads will be interrupted.

[0207] If the load shedding needs of the main grid are still not met, the distribution network will be fed back to the main station control center to increase the load shedding of other distribution networks while ensuring the safety and stability of the power grid risk recovery. The objective function of the distribution network optimal load shedding model is:

[0208] ; (55)

[0209] in, is the number of feeders; A feeder or feeder combination in the distribution network; The load for the selected feeder or feeder combination; is the importance of the selected feeder or feeder combination, which is determined by the important load and the proportion of distributed photovoltaic on the feeder. The constraints are:

[0210] ; (56)

[0211] in, is the total load that needs to be cut.

[0212] The main distribution layered emergency control strategy in the present invention serves as the control layer of the entire solution. When the voltage characterization index monitors an unstable trend or a fault occurs, it quickly realizes the multi-type resource scheduling of the system. The control strategy is divided into a main control unit and a distribution area unit. The main control unit formulates a global emergency control strategy based on the characterization and evaluation results, while the distribution area unit coordinates various types of resources within the region to achieve hierarchical and zoned control. The preliminary preparations for layout optimization and stability evaluation make the main distribution layered control more accurate and efficient in response speed and scheduling effect. When a large fluctuation in grid frequency or voltage is detected, the backup output support provided by the energy storage and photovoltaic system is first called to give priority to ensuring the stability of frequency and voltage; when the fault is further aggravated, the zone load removal measures are gradually implemented to quickly restore the system to a stable operating state. This emergency control strategy further strengthens the initiative of response and the flexibility of control on the basis of existing power system technology. Through the coordinated control of multiple areas, it effectively responds to the volatility of the system after large-scale photovoltaic access, and enhances the stability and recovery capability of the power grid.

[0213] A novel coordinated control method for multiple types of resources in a power system proposed in Example 1 of the present invention is a coordinated control method for multiple resources based on the coordinated layout of energy storage and photovoltaics, voltage stability margin assessment, and hierarchical emergency control. The layout optimization, voltage stability characterization and assessment method, and the main and distribution hierarchical emergency control system are three-layer progressive and interdependent system solutions. Layout optimization ensures the rationality of resource distribution, provides a physical basis for voltage stability, and then achieves real-time monitoring of system status through characterization and assessment, while the main and distribution hierarchical control quickly dispatches multiple types of resources according to the real-time status when necessary, ultimately ensuring the stable operation of the system.

[0214] Example 2

[0215] Based on a novel electric power system multi-type resource coordinated control method proposed in Example 1 of the present invention, Example 2 of the present invention further proposes a novel electric power system multi-type resource coordinated control system, Figure 6 This is a schematic diagram of a novel power system multi-type resource coordination control system proposed in Embodiment 2 of the present invention, the system comprising: an optimization layout module, a calculation module and a coordination control module;

[0216] The optimization layout module is used to establish the power output model of BESS and determine the voltage stability index for transient voltage stability research; the cross entropy is used to optimize the voltage stability index to determine the node set of the top emergency situations to complete the distribution layout of the energy storage unit; the state vector of the combination of photovoltaic power generation prediction and battery storage level is defined in the MDP model, and after the state vector is discretized, different types of costs are determined in the MDP model, and the total cost is calculated to complete the layout optimization of energy storage and photovoltaic units to enhance the system voltage stability;

[0217] The calculation module is used to calculate the voltage stability index to capture the voltage fluctuation and safety margin in the system by using the approximate characteristics of the Jacobian determinant of the power flow after completing the optimization layout of the energy storage and photovoltaic units for voltage stability;

[0218] The coordinated control module is used to determine whether the new power system needs emergency control based on the voltage stability index. When control is needed, the energy storage working state is classified according to the shortest discharge time and the remaining discharge time, and the output power control under different working states is determined; the load reduction parameters are calculated according to the output power control under different states, and the emergency control strategy of the main distribution layer is adopted to realize the coordinated control of multiple types of resources.

[0219] In the optimization layout module: the specific process of establishing the BESS power output model and determining the voltage stability index for transient voltage stability research includes:

[0220] The power output model of BESS is established as:

[0221] ; (1)

[0222] ; (2)

[0223] in, is the power output of BESS; is the time constant of the BESS power converter; represents the reference power; represents the time step; Represents the current state of charge of the BESS; Represents the maximum value of the BESS state of charge; Represents the minimum value of the BESS state of charge; Represents the maximum limit of the power output of the BESS; is the controller gain of each BESS; Represents the frequency change of the system;

[0224] The specific process of determining the voltage stability indicators used in transient voltage stability studies includes:

[0225] Assuming there is a previous Serious failures need to be handled, for node indicators ,exist The percentage change of voltage at time is:

[0226] ; (3)

[0227] in, For the The percentage of voltage change after a fault occurs; Node index exist Failure Voltage at the moment; Node index Initial voltage when no fault occurs; is the total number of node indicators; is the node indicator; is the maximum number of time steps;

[0228] ; (4)

[0229] in, For the The severity index of each fault; For Node exist Fault time Severity index at the time of is the total number of nodes;

[0230] The maximum voltage recovery sensitivity is used for voltage stability index calculation:

[0231] ; (5)

[0232] in, is the sensitivity of the voltage recovery degree to the theoretical maximum power output of all placed BESS units; For Node exist The voltage value before the fault; For Node exist The voltage value after a fault; For all BESS power converters ;

[0233] No. Failure No. The overall voltage stability index at each node is:

[0234] ; (6)

[0235] The overall voltage stability index for all fault configuration combinations is:

[0236] ; (7)

[0237] Assume that there are a total of BESS units are placed to optimize the voltage stability index, then the optimal placement problem of multiple BESS units is:

[0238] ; (8)

[0239] ; (9)

[0240] in, is a binary variable, The number of BESS units to be placed to achieve the optimal voltage stability index; Represents the voltage stability margin.

[0241] The process of optimizing the voltage stability index by cross entropy to determine the top-ranked node set of emergency situations and completing the distribution layout of energy storage units includes: extracting samples based on probability distribution; and updating the parameters of the probability distribution of the extracted samples based on the extracted samples.

[0242] The process of defining the state vector of the combination of PV generation forecast and battery storage level in the MDP model is:

[0243] ; (20)

[0244] in, is the relative error of the PV power generation prediction of the first DER; For the The relative error of the PV power generation forecast for each DER; is the battery storage level of the first BESS; For the The battery storage level of each BESS; is the number of DERs; is the number of BESS units;

[0245] The relative error of the PV power generation forecast for each DER is defined as:

[0246] ;(twenty one)

[0247] in, is the current output level of the photovoltaic unit; is the reference power output of the unit; is the relative error of the PV power generation prediction for each DER;

[0248] The system action can be expressed as: ;(twenty two)

[0249] in, is the action vector; is the reduction in PV power generation of the first DER; For the PV generation reduction per DER; The injected reactive power of the first DER; For the The injected reactive power of each DER; Charge and discharge the DER battery of the first BESS; For the The DER batteries of each BESS are charged and discharged; is the load shedding amount;

[0250] Based on historical data, a to another The transition probability of the state;

[0251] ;(twenty three)

[0252] in, For photovoltaic units from the state arrive The probability of For the first state; For the second state; is the number of times the transfer occurred; for The total number of times the state occurs;

[0253] Each unit and BESS The transition probabilities are independent of each other, so the obtained transition probability is equal to the product of all probabilities; accordingly, from the third state To the fourth state In action The transition probability under is as follows:

[0254] ;(twenty four)

[0255] in Represents the energy storage unit from the original state Transfer to state The probability of Represents the photovoltaic unit from the original state Transfer to state The probability of By Status To status In action The transition probability under .

[0256] The process of determining different types of costs in the MDP model and calculating the total cost to complete the optimal layout of energy storage and photovoltaic units to enhance system voltage stability includes:

[0257] The total cost in time is:

[0258] ; (25)

[0259] in, represent Total cost of time; represents the equivalent cost of underutilization of PV resources; Represents the value of abandoned electricity; represents a penalty for load shedding; Represents the load shedding value; Represents the load shedding loss on the power grid side; Represents the load loss on the side of distributed energy resources; Represents the equivalent network loss cost of the power grid; represents the equivalent power loss cost of distributed energy resources;

[0260] for Flexibility of services provided; For active power transfer between distribution and transmission systems; For reactive power transfer between distribution and transmission systems;

[0261] The method of the offline calculation stage is:

[0262] ; (26)

[0263] in, Calculate cost values ​​for each time period, state, and measure; Represents offline computing Expected utility at a given moment; represents the penalty coefficient, which represents the penalty cost after violating the constraint conditions; represent At any time, the state To status In action The transition probability under ; Represents the expected utility in the next period; Represents the possible state set of offline computing presets;

[0264] The method of the online calculation phase is:

[0265] ; (27)

[0266] in, Represents the instantaneous cost during online computing; Represents the actual set of actions that may occur during online computing; represents the immediate cost for each action; represent At any time, the state To status In action The transition probability under ;

[0267] At this point, the optimized layout of energy storage and photovoltaic units to enhance system voltage stability has been completed.

[0268] In the calculation module, the process of constructing a voltage stability index to capture the voltage fluctuation and safety margin in the system by using the approximate characteristics of the power flow Jacobian determinant includes:

[0269] Modeling distribution networks using relaxed branch power flow equations:

[0270] ; (28)

[0271] in, Represents the active power of the node; Representative Node The total active power to the adjacent nodes; Representative Node resistance to adjacent nodes; Representative Node The square of the complex current amplitude to the adjacent node; let Represents the set of all its child nodes; Representative branch Active power; Represents the node reactive power; Representative Node The sum of reactive power to adjacent nodes; Representative Node reactance to adjacent nodes; Representative branch Reactive power; Representative Node The square of the complex voltage amplitude at ; Representative Node The square of the complex voltage amplitude at ; Representative branch resistance; Representative branch Active power; Representative branch Reactance; Representative branch Reactive power; Representative Node To Node The square of the branch complex current amplitude;

[0272] The relaxed vector form of the branch flow equation is:

[0273] ; (29)

[0274] in, , and , is the defined phasor, respectively, by the scalar , and , Stacking results; and Represents the two (0, 1) matrices defined in this paper, where For each line Select branch The matrix of For each line Select branch Matrix of represent Eliminate The matrix obtained by row; represent The transposed matrix of Represents an n×n matrix, with and The phasors obtained by multiplication are diagonal elements and zero elsewhere; Represents the resistance matrix, which is an n×n matrix, and the resistance on the diagonal and zero elsewhere; Represents the reactance matrix, which is an n×n matrix. on the diagonal and zero elsewhere; Represents the branch active power matrix, which is an n×n matrix. on the diagonal and zero elsewhere; Represents the branch reactive power matrix, which is an n×n matrix. on the diagonal and zero elsewhere; represents the square of the complex current amplitude; represents the square of the complex voltage amplitude;

[0275] The Jacobian matrix of the power flow equation is formalized to represent the standard reasoning for the load capacity limit and the result is used for the adopted branch power flow model, considering two vectors:

[0276] ; (30)

[0277] in, and Represents two vectors, corresponding to system variables and system parameters respectively; and Represents the active and reactive power flows of each branch; Represents the square of the branch complex current amplitude; represents the square of the node complex voltage amplitude; and Represents the active and reactive power of the node load; Represents the system voltage; Represents the system active power flow; Represents the reactive power flow of the system; the reactive power flow is the trend of reactive current flowing from the power source into the loads of each component of the system; represents the first spatial set; represents the second space set;

[0278] Formula (29) is expressed in implicit form as:

[0279] ; (31)

[0280] The loadability limit corresponds to the scalar function The maximum value of the physical grid constraint is the set ,

[0281] ; (32)

[0282] The relaxed branch power flow model describes a network with 1 relaxed node and The voltage stability domain of the distribution network with a PQ node is an open area where the power flow solution set satisfies:

[0283] ; (33)

[0284] When the branching flow model is adopted, Take the form:

[0285] ; (34)

[0286] in, and They are respectively and The matrix obtained by removing the first row is and They are and The transposed matrix of is the first canonical basis vector; Represents a The all-zero area means that all elements in this area are 0;

[0287] Formula (34) is simplified to the power flow Jacobian matrix:

[0288] ; (35)

[0289] in, represents the reduced-order power flow Jacobian;

[0290] use As an indicator of voltage stability;

[0291] ; (36)

[0292] in, It is the actual voltage stability index;

[0293] For a certain threshold , the actual voltage stability measure becomes:

[0294] ; (37)

[0295] in, is the minimum value of the voltage stability index;

[0296] The voltage stability index is approximately defined as:

[0297] ; (38)

[0298] in, It is the approximate value of voltage stability index; for The product of the elements on the diagonal;

[0299] The approximate interval of the voltage stability index is:

[0300] ; (39).

[0301] In the coordinated control module, the energy storage working state is classified according to the shortest discharge time and the remaining discharge time, including: Longer than the shortest discharge time When the energy storage capacity is sufficient, it is the first working state; the remaining discharge time The duration of energy storage at maximum power output; the shortest discharge time Refers to the shortest time that the grid requires energy storage to provide power support in an emergency; the remaining discharge time Shorter than the shortest discharge time When the energy storage capacity is insufficient, if it is in the charging state, it is the second working state of the energy storage; when it is in the standby or discharging state, if the capacity is insufficient but power support can be provided, it is the third working state; if the capacity is lower than that which makes it difficult to provide power support, it is the fourth working state, and additional control is added to stop charging and discharging.

[0302] Determine the output power control under different working conditions specifically including:

[0303] For the energy storage in the first state, the power shortage is supported by discharging at the maximum output within the system required time;

[0304] ; (40)

[0305] in, Energy storage for the first working state Output power when emergency power support is needed; for The remaining energy storage capacity at any time; The minimum allowable remaining power to provide emergency support; is the energy storage discharge efficiency;

[0306] The second working state in the charging state is converted into the third working state or the fourth working state according to the charging state and the delay requirement of the unit;

[0307] In the third working state, the power output is reduced to achieve the shortest discharge time, and the output power control is:

[0308] ; (41)

[0309] in, Energy storage for the third working state Output power when emergency power support is needed; The charge that can be maintained for the shortest power supply time under the maximum discharge state; is the maximum amount of energy stored;

[0310] In the fourth working state, no control is applied.

[0311] The process of calculating the load reduction parameters according to the output power control in different states includes: calculating the frequency change rate of the system; calculating the minimum peak occurrence time after the system has a power shortage according to the frequency change rate of the system; calculating the maximum frequency deviation, the minimum frequency and the stabilized frequency according to the minimum peak occurrence time; through the improved system frequency response model, combined with the initial frequency change rate of each node monitored by the wide area measurement system, the actual power shortage caused by the system failure is calculated, and then the load shedding amount of the unit's primary frequency regulation is calculated.

[0312] A new type of coordinated control system for multiple types of resources in a power system proposed in Example 2 of the present invention is a coordinated control method for multiple resources based on the coordinated layout of energy storage and photovoltaics, voltage stability margin assessment, and hierarchical emergency control. The layout optimization, voltage stability characterization and assessment method, and the main and distribution hierarchical emergency control system are three-layer progressive and interdependent system solutions. Layout optimization ensures the rationality of resource distribution, provides a physical basis for voltage stability, and then achieves real-time monitoring of system status through characterization and assessment, while the main and distribution hierarchical control quickly dispatches multiple types of resources according to the real-time status when necessary, ultimately ensuring the stable operation of the system.

[0313] The description of the relevant parts of a new type of power system multi-type resource coordination control system provided in Example 2 of the present application can be found in the detailed description of the corresponding parts of a new type of power system multi-type resource coordination control method provided in Example 1 of the present application, and will not be repeated here.

[0314] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply that there is any such actual relationship or order between these entities or operations. Moreover, the term "include", "comprise" or any other variant thereof is intended to cover non-exclusive inclusion, so that the process, method, article or equipment that includes a series of elements are inherent to the elements. In the absence of more restrictions, the elements limited by the sentence "comprise one..." do not exclude the presence of other identical elements in the process, method, article or equipment that includes the elements. In addition, the above-mentioned technical solution provided in the embodiment of the present application is consistent with the corresponding technical solution in the prior art in principle, and the part is not described in detail, so as not to repeat too much.

[0315] Although the above describes the specific implementation of the present invention in conjunction with the accompanying drawings, it is not intended to limit the scope of protection of the present invention. For those skilled in the art, other different forms of modifications or deformations can be made on the basis of the above description. It is not necessary and impossible to list all the implementation methods here. On the basis of the technical solution of the present invention, various modifications or deformations that can be made by those skilled in the art without creative work are still within the scope of protection of the present invention.

Claims

1. A novel coordinated control method for multiple types of resources in a power system, characterized in that: The following steps are involved: Establish a power output model of the battery energy storage system BESS and determine the voltage stability index for transient voltage stability research; use cross entropy to optimize the voltage stability index to determine the top-ranked emergency node set to complete the distribution layout of the energy storage unit; define the state vector of the combination of photovoltaic power generation prediction and battery storage level in the MDP model, discretize the state vector, determine different types of costs in the MDP model, and calculate the total cost to complete the optimization layout of energy storage and photovoltaic units to enhance system voltage stability; After completing the optimal layout of energy storage and photovoltaic units for voltage stability, the voltage stability index is constructed using the approximate characteristics of the Jacobian determinant of the power flow to capture the voltage fluctuation and safety margin in the system; According to the voltage stability index, it is judged whether the new power system needs emergency control. When control is needed, the energy storage working state is classified according to the shortest discharge time and the remaining discharge time, and the output power control under different working states is determined; the load reduction parameters are calculated according to the output power control under different states, and the emergency control strategy of main distribution layering is adopted to realize the coordinated control of multiple types of resources; The process of determining different types of costs in the MDP model, calculating the total cost, and completing the optimal layout of energy storage and photovoltaic units to enhance system voltage stability includes: The total cost in time is: ； in, represent Total cost of time; represents the equivalent cost of underutilization of PV resources; Represents the value of abandoned electricity; represents a penalty for load shedding; Represents the load shedding value; Represents the load shedding loss on the power grid side; Represents the load loss on the side of distributed energy resources; Represents the equivalent network loss cost of the power grid; represents the equivalent power loss cost of distributed energy resources; for Flexibility of services provided; For active power transfer between distribution and transmission systems; For reactive power transfer between distribution and transmission systems; The method of the offline calculation stage is: ； in, Calculate cost values ​​for each time period, state, and measure; Represents offline computing Expected utility at a given moment; represents the penalty coefficient, which represents the penalty cost after violating the constraint conditions; represent At any time, the state To status In action The transition probability under ; Represents the expected utility in the next period; Represents the set of states that exist in offline computing presets; The method of the online calculation phase is: ； in, Represents the instantaneous cost during online computing; Represents the set of actions that actually occur during online computing; represents the immediate cost for each action; represent At any time, the state To status In action The transition probability under ; At this point, the optimal layout of energy storage and photovoltaic units to enhance system voltage stability has been completed; The process of constructing a voltage stability index to capture voltage fluctuations and safety margins in the system using the approximate characteristics of the power flow Jacobian determinant includes: Modeling distribution networks using relaxed branch power flow equations: ； in, Represents the active power of the node; Representative Node The total active power to the adjacent nodes; Representative Node resistance to adjacent nodes; Representative Node The square of the complex current amplitude to the adjacent node; let Represents the set of all its child nodes; Representative branch Active power; Represents the node reactive power; Representative Node The sum of reactive power to adjacent nodes; Representative Node reactance to adjacent nodes; Representative branch Reactive power; Representative Node The square of the complex voltage amplitude at ; Representative Node The square of the complex voltage amplitude at ; Representative branch resistance; Representative branch Active power; Representative branch Reactance; Representative branch Reactive power; Representative Node To Node The square of the branch complex current amplitude; The relaxed vector form of the branch flow equation is: ； in, , and , is the defined phasor, respectively, by the scalar , and , Stacking results; and Represents the two (0, 1) matrices defined in this paper, where For each line Select branch The matrix of For each line Select branch Matrix of represent Eliminate The matrix obtained by row; represent The transposed matrix of Represents an n×n matrix, with and The phasors obtained by multiplication are diagonal elements and zero elsewhere; Represents the resistance matrix, which is an n×n matrix, and the resistance on the diagonal and zero elsewhere; Represents the reactance matrix, which is an n×n matrix. on the diagonal and zero elsewhere; Represents the branch active power matrix, which is an n×n matrix. on the diagonal and zero elsewhere; Represents the branch reactive power matrix, which is an n×n matrix. on the diagonal and zero elsewhere; represents the square of the complex current amplitude; represents the square of the complex voltage amplitude; The Jacobian matrix of the power flow equation is formalized to represent the standard reasoning for the load capacity limit and the result is used for the adopted branch power flow model, considering two vectors: ； in, and Represents two vectors, corresponding to system variables and system parameters respectively; and Represents the active and reactive power flows of each branch; Represents the square of the branch complex current amplitude; represents the square of the node complex voltage amplitude; and Represents the active and reactive power of the node load; Represents the system voltage; Represents the system active power flow; Represents the system reactive power flow; represents the first spatial set; represents the second space set; The relaxed vector form of the branch flow equation is expressed in implicit form as: ； The loadability limit corresponds to the scalar function The maximum value of the physical grid constraint is the set , ； The relaxed branch power flow model describes a network with 1 relaxed node and The voltage stability domain of the distribution network with a PQ node is an open area where the power flow solution set satisfies: ； When the branching flow model is adopted, Take the form: ； in, and They are respectively and The matrix obtained by removing the first row is and They are and The transposed matrix of is the first canonical basis vector; Represents a The all-zero area means that all elements in this area are 0; The branch flow model is simplified to the power flow Jacobian matrix: ； in, represents the reduced-order power flow Jacobian; use As an indicator of voltage stability; ； in, It is the actual voltage stability index; For a certain threshold , the actual voltage stability measure becomes: ； in, is the minimum value of the voltage stability index; The voltage stability index is approximately defined as: ； in, It is the approximate value of voltage stability index; for The product of the elements on the diagonal; The approximate interval of the voltage stability index is: 。 2. A novel method for coordinated control of multiple types of resources in a power system according to claim 1, characterized in that: The specific process of establishing the power output model of the battery energy storage system BESS and determining the voltage stability index for transient voltage stability research includes: The power output model of BESS is established as: ； ； in, is the power output of BESS; is the time constant of the BESS power converter; represents the reference power; represents the time step; Represents the current state of charge of the BESS; Represents the maximum value of the BESS state of charge; Represents the minimum value of the BESS state of charge; Represents the maximum limit of the power output of the BESS; is the controller gain of each BESS; Represents the frequency change of the system; The specific process of determining the voltage stability indicators used in transient voltage stability studies includes: Assuming there is a previous Serious failures need to be handled, for node indicators ,exist The percentage change of voltage at time is: ； in, For the The percentage of voltage change after a fault occurs; Node index exist Failure Voltage at the moment; Node index Initial voltage when no fault occurs; is the total number of node indicators; is the node indicator; is the maximum number of time steps; ； in, For the The severity index of each fault; For Node exist Fault time Severity index at the time of is the total number of nodes; The maximum voltage recovery sensitivity is used for voltage stability index calculation: ； in, is the sensitivity of the voltage recovery degree to the theoretical maximum power output of all placed BESS units; For Node exist The voltage value before the fault; For Node exist The voltage value after a fault; For all BESS power converters ; No. Failure No. The overall voltage stability index at each node is: ； The overall voltage stability index for all fault configuration combinations is: ； Assume that there are a total of BESS units are placed to optimize the voltage stability index, then the optimal placement problem of multiple BESS units is: ； ； in, is a binary variable, The number of BESS units to be placed to achieve the optimal voltage stability index; Represents the voltage stability margin.

3. A novel method for coordinated control of multiple types of resources in a power system according to claim 2, characterized in that: The process of using cross entropy to optimize the voltage stability index and determine the top-ranked emergency node set to complete the distribution layout of the energy storage unit includes: Draw samples based on probability distribution; Based on the drawn samples, update the parameters of the probability distribution of the drawn samples.

4. A novel method for coordinated control of multiple types of resources in a power system according to claim 1, characterized in that: The process of defining the state vector of the combination of PV generation forecast and battery storage level in the MDP model is: ； in, is the relative error of the PV power generation prediction of the first DER; For the The relative error of the PV power generation forecast for each DER; is the battery storage level of the first BESS; For the The battery storage level of each BESS; is the number of DERs; is the number of BESS units; The relative error of the PV power generation forecast for each DER is defined as: ； in, is the current output level of the photovoltaic unit; is the reference power output of the unit; is the relative error of the PV power generation prediction for each DER; The system action is expressed as: ; in, is the action vector; is the reduction in PV power generation of the first DER; For the PV generation reduction per DER; The injected reactive power of the first DER; For the The injected reactive power of each DER; Charge and discharge the DER battery of the first BESS; For the The DER battery of each BESS charges and discharges; is the load shedding amount; Based on historical data, a to another The transition probability of the state; ； in, For photovoltaic units from the state arrive The probability of For the first state; For the second state; is the number of times the transfer occurred; for The total number of times the state occurs; Each unit and BESS The transition probabilities are independent of each other, so the obtained transition probability is equal to the product of all probabilities; accordingly, from the third state To the fourth state In action The transition probability under is as follows: ； in Represents the energy storage unit from the original state Transfer to state The probability of Represents the photovoltaic unit from the original state Transfer to state The probability of By Status To status In action The transition probability under .

5. A novel method for coordinated control of multiple types of resources in a power system according to claim 1, characterized in that: The classification of the energy storage working state according to the shortest discharge time and the remaining discharge time specifically includes: Remaining discharge time Longer than the shortest discharge time When the energy storage capacity is sufficient, it is the first working state; the remaining discharge time The duration of energy storage at maximum power output; the shortest discharge time Refers to the shortest time that the grid requires energy storage to provide power support in an emergency; Remaining discharge time Shorter than the shortest discharge time When the energy storage capacity is insufficient, if it is in the charging state, it is the second working state of the energy storage; In standby or discharge state, if the power is insufficient but sufficient for power support, it is the third working state; If the battery level is lower than the level that makes it difficult to support power, the fourth working state is entered, and the control amount is added to stop charging and discharging.

6. A novel method for coordinated control of multiple types of resources in a power system according to claim 5, characterized in that: The determining of output power control under different working states specifically includes: For the energy storage in the first state, the power shortage is supported by discharging at the maximum output within the system required time; ； in, Energy storage for the first working state Output power when emergency power support is needed; for The remaining energy storage capacity at any time; The minimum allowable remaining power to provide emergency support; is the energy storage discharge efficiency; The second working state in the charging state is converted into the third working state or the fourth working state according to the charging state and the delay requirement of the unit; In the third working state, the power output is reduced to achieve the shortest discharge time, and the output power control is: ； in, Energy storage for the third working state Output power when emergency power support is needed; The charge corresponding to the shortest power supply time requirement under the maximum discharge state; is the maximum amount of energy stored; In the fourth working state, no control is applied.

7. A novel coordinated control method for multiple types of resources in a power system according to claim 5, characterized in that: The process of calculating the load shedding parameters according to the output power control in different states includes: Calculate the frequency change rate of the system; According to the frequency change rate of the system, the minimum peak time after the system has power shortage is calculated; Calculate the maximum frequency deviation, the minimum frequency and the stable frequency according to the minimum peak occurrence time; Through the improved system frequency response model, combined with the initial frequency change rate of each node monitored by the wide-area measurement system, the actual power shortage caused by the system failure is calculated, and then the load shedding amount of the unit's primary frequency regulation is calculated.

8. A novel electric power system multi-type resource coordination control system, used to execute a novel electric power system multi-type resource coordination control method according to any one of claims 1 to 7, characterized in that: include: Optimize the layout module, calculation module and coordination control module; The optimization layout module is used to establish a power output model of the battery energy storage system BESS and determine the voltage stability index for transient voltage stability research; the cross entropy is used to optimize the voltage stability index to determine the node set of the top emergency situations to complete the distribution layout of the energy storage unit; the state vector of the combination of photovoltaic power generation prediction and battery storage level is defined in the MDP model, and after the state vector is discretized, different types of costs are determined in the MDP model, and the total cost is calculated to complete the layout optimization of energy storage and photovoltaic units to enhance system voltage stability; The calculation module is used to calculate the voltage stability index for capturing the voltage fluctuation and safety margin in the system by using the approximate characteristics of the Jacobian determinant of the power flow after completing the energy storage and photovoltaic unit layout optimization arrangement for voltage stability; The coordinated control module determines whether the new power system needs emergency control based on the voltage stability index. When control is needed, the energy storage working state is classified according to the shortest discharge time and the remaining discharge time, and the output power control under different working states is determined; the load reduction parameters are calculated according to the output power control under different states, and a main distribution layered emergency control strategy is adopted to realize coordinated control of multiple types of resources.

Images