Power system reliability constraint economic dispatching method based on critical state

Abstract

The invention discloses a power system reliability constraint economic dispatching method based on a critical state, and the method comprises the steps: screening a system state according to the LOLP index requirement of a power system, removing a state which must lose a load through a chance constraint DC optimal power flow model, and screening a worse state in a state set according to the capacity of a fault element, and finally, in a source load prediction error range, according to the tolerance of the system state to the output plan, screening out worse states in the set so as to obtain a critical state set. According to the method, the reliability requirement is converted into the constraint condition that no load loss occurs in the critical state, the state number is effectively reduced, the calculation complexity of the reliability constraint economic dispatching of the power system is reduced, and the efficiency and accuracy of the reliability constraint economic dispatching of the power system are remarkably improved.

Description

Technical Field

[0001] The present invention relates to the technical field of power system dispatching, and in particular to a power system reliability-constrained economic dispatching method based on critical state. Background Art

[0002] In the past few decades, with the rapid growth of electricity demand and the large-scale grid-connected operation of wind power, the safe and reliable operation of the power system has been affected. There are roughly two factors that affect the reliability of power system operation. On the one hand, it comes from the intermittent and uncertain nature of wind power and the random fluctuations of load. On the other hand, it is mainly caused by the random outage of components. In view of this, it is necessary to solve the economic dispatch with operation risks or reliability issues to improve the reliability of power system operation.

[0003] However, in the study of component failure, reliability-constrained economic scheduling can be divided into safety-constrained economic scheduling and risk-constrained scheduling according to whether the probability of event occurrence is considered. Safety-constrained economic scheduling requires that the system can still operate reliably after the expected system state occurs. Risk-constrained scheduling expresses reliability indicators as objective functions or constraints, and controls the risk of the expected system state rather than individual accidental events. The complexity of risk scheduling and safety-constrained economic scheduling increases with the increase of the expected system state. Since the number of system states grows exponentially, the computational complexity of the economic scheduling model is extremely high. Therefore, reducing the system state is an effective means to establish a reliability-constrained economic scheduling model. The reduction process of the expected system state determines the economy and reliability of the scheduling decision.

[0004] In the prior art, several methods have been proposed, such as umbrella contingencies, universal generating functions, uniform design, and root events to reduce the computational burden by reducing system states. However, existing methods for reducing expected system states focus on screening high-risk system states to control expected short supply (EENS) in reliability-constrained economic dispatch, and it is difficult to consider the loss of load probability (LOLP).

[0005] Therefore, how to propose a state screening method to significantly reduce the number of states used for reliability constraints while maintaining the high accuracy of reliability indicators is a technical problem that currently needs to be solved urgently by technical personnel in this field. Summary of the invention

[0006] In view of the deficiencies of the above-mentioned prior art, the present invention provides a reliability-constrained economic dispatch method for power systems based on critical states, which converts the reliability requirements into constraint conditions that no load loss occurs in critical states, so as to reduce the computational complexity of reliability-constrained economic dispatch of power systems and improve the operational reliability of power systems.

[0007] In order to solve the above technical problems, the present invention adopts the following technical solutions:

[0008] A reliability-constrained economic dispatch method for a power system based on criticality comprises the following steps:

[0009] S1. Obtain the LOLP index requirement of the power system, and obtain the cumulative probability requirement of the truncation state according to the obtained LOLP index requirement of the power system;

[0010] S2. According to the cumulative probability requirement of the truncated state, the system state is selected based on the descending order of probability to obtain the system state set H;

[0011] S3. According to the acquired system state set, a chance-constrained DC optimal power flow model is used to obtain a corresponding load loss amount to obtain a system state without load loss, thereby obtaining a state set I;

[0012] S4, taking the upper and lower limits of each bus conventional unit of the system state in the calculated state set I and the upper and lower limits of the transmission capacity between each bus pair as the criterion of the good-bad relationship between the system states in the state set I, updating the state set I so that the system state in the state set I is the worst state, thereby obtaining the state set A;

[0013] S5, dividing the system states in the state set A into different subsets according to whether the faulty units are the same, and screening each subset in turn to select the worst system state, thereby obtaining the critical state set C;

[0014] S6. Based on the critical state set C, a reliability-constrained economic dispatch model with reliability constraints is constructed and solved to obtain a dispatch plan that meets the LOLP requirements.

[0015] Preferably, in step S1, the cumulative probability of the truncation state is required to be calculated by the following formula:

[0016]

[0017] In the formula, represents the cumulative probability requirement of the truncated state, represents the LOLP index requirement of the power system, and ΔP represents the slack variable that makes the screening state meet the LOLP index requirement of the power system.

[0018] Preferably, step S2 comprises:

[0019] S21, initialize the cumulative probability P = 0, the currently selected system state set H is an empty set, and the currently selected maximum state order k = 0;

[0020] S22, calculating the occurrence probability of each system state in the system state set, and arranging the occurrence probability of each system state in descending order Where S i,m represents the mth selected system state in the i-th order state, The probability of occurrence of the mth selected system state in the i-th order state;

[0021] S23, according to the sorted system state list, select the system state ranked first in the system state list and add it to the system state set H, and update the cumulative probability of the current selection

[0022] S24, when it is determined that the cumulative probability of the currently selected system state is greater than the cumulative probability of the truncated state, executing step S25; otherwise, executing step S26;

[0023] S25, the system state selection stops, and the neighborhood state of the selected system state is added to the system state set H, and the final system state H is output;

[0024] S26. Take the next state in the system state list as the currently selected system state, and return to step S23.

[0025] Preferably, in step S22, the calculation formula for calculating the occurrence probability of each system state in the system state set is as follows:

[0026]

[0027] Where P(s) represents the probability of a single system state s, X represents the set of all components in the power system, and X d represents the set of faulty components in the system state, p i represents the failure probability of component i.

[0028] Preferably, in step S3, a chance-constrained DC optimal power flow model is used to obtain a corresponding calculation formula for the load loss amount:

[0029]

[0030] In the formula, p g , They represent the minimum output, planned output and maximum output of the conventional unit g, respectively, and p w 、p w,f , They represent the planned output, predicted power generation and actual power generation of wind turbine w respectively; d b ,d b,f , They represent the planned supply value, predicted value, and actual value of the load on bus b, respectively. a represents the total deviation of injected power caused by source charge prediction error, They represent the power distribution factors of the conventional unit injection power, renewable energy injection power, and load demand converted into line power flow, respectively. represents the abandoned air volume of the wind turbine w, f l max represents the upper limit of the transmission capacity of line l, ε UR , ε DR , ε L They represent the confidence level of not violating the upper reserve / lower reserve / line power flow constraints respectively, G represents the set of all conventional units, W represents the set of all wind turbines, and B represents the set of all buses.

[0031] Preferably, in step S4, the calculation formula for the upper and lower limits of each bus conventional unit in the system state in the state set I is as follows:

[0032]

[0033] In the formula, represents the output upper limit of the conventional units on bus b under system state s, Indicates the lower limit of the output of the conventional unit on bus b under system state s, Indicates the state of conventional unit g under system state s;

[0034] The calculation formula for the upper and lower limits of the transmission capacity between the bus pairs is as follows:

[0035]

[0036] In the formula, represents the upper limit of the transmission capacity between the bus pairs (k, b) in the system state s, represents the lower limit of the transmission capacity between the bus pairs (k, b) in the system state s, represents the state of line l in state s, and (k, b) represents a bus pair.

[0037] Preferably, step S5 comprises:

[0038] S51, initializing the critical state set C to be an empty set;

[0039] S52. Setting different subsets N g Indicates the number of types of faulty units, and sets The first state in Set to the baseline state and initialize the data used to store the data from the sub-collection The state set filtered out is an empty set;

[0040] S53, calculating the reference state Under the condition of no loss of load, the collection The maximum optimal load shedding LS c and the corresponding state c;

[0041] S54, when LS c When LS is equal to 0, step S55 is executed; c If it is greater than 0, update for renew for Return to step S53;

[0042] S55, will Updated to As the new baseline state, until the subset When each state in is taken as the reference state, the critical state set C is updated as

[0043] Preferably, in step S53, the calculation of the reference state Under the condition of no loss of load, the collection The maximum optimal load shedding LS c And the corresponding state c is calculated as follows:

[0044] maxLS c

[0045]

[0046] In the formula, p w,f They represent the actual value and predicted value of wind turbine w power generation, d b,f They represent the actual value and predicted value of busbar b load respectively. They represent the planned output of conventional unit g and wind turbine unit w under state s, respectively, and f l s represents the power flow of line l in state s, The state represents the phase angle of bus bar b in state s, represents the maximum value of the prediction error of wind turbine w, Indicates the maximum value of the load prediction error of bus b.

[0047] Preferably, in step S6, the reliability constraint is to impose a load-free constraint on each state in the critical state set C, and the specific steps are as follows:

[0048]

[0049] In the formula, represents the output fluctuation of conventional unit g caused by component failure in state s, represents the total deviation of injected power caused by source load prediction error in state s, represents the participation factor of the radiation control of the conventional unit g under state s, They represent upper standby and lower standby of conventional units respectively. They respectively represent the power flow distribution factors converted from conventional unit injection power / renewable energy injection power / load demand into line power flow under state s. p w They represent the actual power generation value and planned output of wind turbine w, Indicates the actual value of busbar b load, p g represents the planned output of conventional units g, represents the abandoned air volume of wind turbine w in state s, f l max represents the upper limit of the transmission capacity of line l, represents the state of unit g under state s, ε UR , ε DR , ε L They represent the confidence level of not violating the upper reserve / lower reserve / line power flow constraints, G s represents the set of faulty units in state s, G b represents the set of conventional units on bus b, G represents the set of all conventional units, W represents the set of all wind turbines, and B represents the set of all buses.

[0050] Compared with the prior art, the present invention has the following technical effects:

[0051] (1) The present invention reduces the number of states that need to be considered in the economic dispatch model by screening out the critical state set in the power system, thereby reducing the computational complexity of the model. While reducing the number of states, the high accuracy of the reliability index is ensured, so that the reliability of the power system is not affected. The screened critical state set is used to construct the reliability constraints of the system. These constraints can ensure that no load loss occurs during the operation of the power system. Finally, the constructed reliability constraints are incorporated into the reliability-constrained economic dispatch (RECD) model, and the constraints of the model are modified or expanded, so that the dispatch scheme in the dispatch model can achieve economic dispatch while satisfying the reliability.

[0052] (2) The reliability-constrained economic dispatch method for power systems based on critical states designed by the present invention, which is applied to power system dispatching, constructs reliability constraints based on critical state sets, realizes accurate modeling of power system operation reliability, quickly and accurately evaluates power system operation reliability, and can be applied to reliability-constrained economic dispatching to improve system operation reliability. It can play a good supplementary role in the economic dispatching of traditional power systems and is of great significance to planning and dispatching that take reliability into account. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] In order to make the purpose, technical solution and advantages of the invention more clear, the present invention will be further described in detail below with reference to the accompanying drawings, in which:

[0054] Figure 1 It is a critical state schematic diagram in the present invention;

[0055] Figure 2 A flow chart of a reliability-constrained economic dispatch method for a power system based on a critical state disclosed by the present invention;

[0056] Figure 3 Schematic diagram of neighborhood status in an embodiment of the present invention. DETAILED DESCRIPTION

[0057] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. The components of the embodiments of the present invention generally described and shown in the drawings here can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present invention provided in the drawings is not intended to limit the scope of the claimed invention, but only represents selected embodiments of the present invention. Based on the embodiments in the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work belong to the scope of protection of the present invention.

[0058] The present invention will be further described in detail below in conjunction with the accompanying drawings.

[0059] In order to solve the problem of slow calculation speed of reliability-constrained economic dispatch in power systems, the present invention proposes a screening method based on critical states, which converts reliability requirements into constraint conditions that no load loss occurs in critical states. This method takes into account the complexity of reliability constraints, facilitates their inclusion in dispatch models, and provides more reliable support for the reliable and safe operation of new power systems.

[0060] In order to achieve the above objectives, it is necessary to establish the concept of critical state of power system, that is, during the operation of power system, there are usually specific requirements for loss of load probability (LOLP), such as the requirement of "no more than one day of power outage in ten years" can be converted into LOLP requirement. Figure 1 The critical state shown is a part of the specific state. If this part of the state does not lose the load, the state set with the cumulative probability of satisfying 1-LOLP will not lose the load, so that the reliability of the power system meets the requirements. Using this part of the state to establish reliability constraints can greatly reduce the complexity of reliability constraints and facilitate the solution of reliability-constrained economic dispatch.

[0061] In specific implementation, the steps of the present invention are as follows Figure 2 As shown, it specifically includes:

[0062] S1. Obtain the LOLP index requirement of the power system, and obtain the cumulative probability requirement of the truncation state according to the obtained LOLP index requirement of the power system;

[0063] S2. According to the cumulative probability requirement of the truncated state, the system state is selected based on the descending order of probability to obtain the system state set H;

[0064] S3. According to the acquired system state set, a chance-constrained DC optimal power flow model is used to obtain a corresponding load loss amount to obtain a system state without load loss. When it is determined that the cumulative probability of the system state without load loss meets the LOLP index requirement of the power system, a state set I is obtained;

[0065] S4, taking the upper and lower limits of each bus conventional unit of the system state in the calculated state set I and the upper and lower limits of the transmission capacity between each bus pair as the criterion of the good-bad relationship between the system states in the state set I, updating the state set I so that the system state in the state set I is the worst state, thereby obtaining the state set A;

[0066] S5. Divide the system states in the state set A into different sub-sets according to whether the faulty units are the same, and screen each sub-set in turn to select the worst system state, thereby obtaining the critical state set C.

[0067] S6. Based on the critical state set C, a reliability-constrained economic dispatch model with reliability constraints is constructed and solved to obtain a dispatch plan that meets the LOLP requirements.

[0068] In this embodiment, by screening out the critical state set in the power system, the number of states that need to be considered in the economic dispatch model is reduced, thereby reducing the computational complexity of the model. While reducing the number of states, the high accuracy of the reliability index is ensured so that the reliability of the power system is not affected. The screened critical state set is used to construct the reliability constraints of the system. These constraints can ensure that no load loss occurs during the operation of the power system. Finally, the constructed reliability constraints are incorporated into the economic dispatch model, and the constraints of the model are modified or expanded so that the dispatch plan in the dispatch model can achieve economic dispatch while satisfying reliability.

[0069] In specific implementation, in step S1, the cumulative probability of the truncation state is required to be calculated by the following formula:

[0070]

[0071] In the formula, represents the cumulative probability requirement of the truncated state, represents the LOLP index requirement of the power system, and ΔP represents the slack variable that makes the screening state meet the LOLP index requirement of the power system.

[0072] In a specific application embodiment, the components are rearranged in ascending order of availability and numbered 1, 2, ..., n. That is, p1≤p2≤...≤p n ; In the formula, P1 represents the availability of component 1; the faulty component is used to represent the system status, such as S3(1,2,3) represents the failure of components 1, 2, and 3.

[0073] In specific implementation, step S2 includes:

[0074] S21, initialize the cumulative probability P = 0, the currently selected system state set H is an empty set, and the currently selected maximum state order k = 0;

[0075] S22, calculating the occurrence probability of each system state in the system state set, and arranging the occurrence probability of each system state in descending order Where S i,m represents the mth selected system state in the i-th order state, The probability of occurrence of the mth selected system state in the i-th order state;

[0076] S23, according to the sorted system state list, select the system state ranked first in the system state list and add it to the system state set H, and update the cumulative probability of the current selection

[0077] S24. When it is determined that the cumulative probability of the currently selected system state is greater than the cumulative probability of the truncated state, step S25 is executed; otherwise, step S26 is executed.

[0078] S25. The system state selection stops, and the neighborhood states of the selected system state are added to the system state set H, and the final system state H is output.

[0079] S26. The next state in the system state list is used as the currently selected system state, and step S23 is returned.

[0080] Specifically, in step S22, the calculation formula for calculating the occurrence probability of each system state in the system state set is as follows:

[0081]

[0082] In the formula, P(s) represents the probability of a single system state s, X represents the set of all components in the power system, and X d represents the set of faulty components in this system state, and p i represents the failure probability of component i.

[0083] Among them, as Figure 3 shown, the generation of neighborhood states follows the following rules: Assume that the serial number of the xth faulty component is "a", and the number of the (x + 1)th faulty component is "c". If there is a faulty component with the number b = a + 1 < c, then replacing the component "a" with "b" can obtain a neighborhood state of the neighborhood state set.

[0084] Specifically, in step S3, using the chance-constrained DC optimal power flow model, the calculation formula for the corresponding load shedding amount is:

[0085]

[0086] In the formula, p g , respectively represent the minimum output, planned output, and maximum output of the conventional unit g, and p w , p w,f , respectively represent the planned output, predicted power generation, and actual power generation of the wind turbine w; d b , d b,f , respectively represent the planned supply value, predicted value, and actual value of the load of bus b, and e a represents the total deviation of the injection power caused by the source-load prediction error, They represent the power distribution factors of the conventional unit injection power, renewable energy injection power, and load demand converted into line power flow, respectively. represents the abandoned air volume of the wind turbine w, f l max represents the upper limit of the transmission capacity of line l, ε UR , ε DR , ε L They represent the confidence level of not violating the upper reserve / lower reserve / line power flow constraints respectively, G represents the set of all conventional units, W represents the set of all wind turbines, and B represents the set of all buses.

[0087] In specific implementation, in step S4, the calculation formula for the upper and lower limits of each bus conventional unit in the system state in the state set I is as follows:

[0088]

[0089] In the formula, represents the output upper limit of the conventional units on bus b under system state s, Indicates the lower limit of the output of the conventional unit on bus b under system state s, Indicates the state of conventional unit g under system state s;

[0090] The calculation formula for the upper and lower limits of the transmission capacity between the bus pairs is as follows:

[0091]

[0092] In the formula, represents the upper limit of the transmission capacity between the bus pairs (k, b) in the system state s, represents the lower limit of the transmission capacity between the bus pairs (k, b) in the system state s, represents the state of line l in state s, and (k, b) represents a bus pair.

[0093] In specific implementation, in step S5, within the source-load prediction error range, under the output of the same conventional unit, the state that is more likely to lose load is a worse state. According to the tolerance of the system state to the output plan, the worse state in the state set A is screened out and used as the critical state set C. When it is ensured that the state in C will not lose load, it can be ensured that the state in the state set I will not lose load. Step S5 includes:

[0094] S51, initializing the critical state set C to be an empty set;

[0095] S52. Setting different subsets N g Indicates the number of types of faulty units, and sets The first state in Set to the baseline state and initialize the data used to store the data from the sub-collection The state set filtered out is an empty set;

[0096] S53, calculating the reference state Under the condition of no loss of load, the collection The maximum optimal load shedding LS c and the corresponding state c;

[0097] S54, when LS c When LS is equal to 0, step S55 is executed; c If it is greater than 0, update for renew for Return to step S53;

[0098] S55, will Updated to As the new baseline state, until the subset When each state in is taken as the reference state, the critical state set C is updated as

[0099] In the specific implementation, in step S53, the calculation of the reference state Under the condition of no loss of load, the collection The maximum optimal load shedding LS c And the corresponding state c is calculated as follows:

[0100] maxLS c

[0101]

[0102] In the formula, p w,f They represent the actual value and predicted value of wind turbine w power generation, d b,f They represent the actual value and predicted value of busbar b load respectively. They represent the planned output of conventional unit g and wind turbine unit w under state s, respectively, and f l s represents the power flow of line l in state s, The state represents the phase angle of bus bar b in state s, represents the maximum value of the prediction error of wind turbine w, Indicates the maximum value of the load prediction error of bus b.

[0103] The constraints that need to be met when performing economic dispatch in the power system include the constraints related to the DC optimal power flow in state c and the reference state; the load shedding in the reference state is 0; the output of the conventional unit in state c is less than or equal to the output of the conventional unit in the reference state; the actual value of the wind turbine power generation and load is within the prediction error range; state c is a set A state in .

[0104] In specific implementation, in step S6, the reliability constraint is to impose a load-free constraint on each state in the critical state set C, and the specific steps are as follows:

[0105]

[0106]

[0107] In the formula, represents the output fluctuation of conventional unit g caused by component failure in state s, represents the total deviation of injected power caused by source load prediction error in state s, represents the participation factor of the radiation control of the conventional unit g under state s, They represent upper standby and lower standby of conventional units respectively. They respectively represent the power flow distribution factors converted from conventional unit injection power / renewable energy injection power / load demand into line power flow under state s. p w They represent the actual power generation value and planned output of wind turbine w, Indicates the actual value of busbar b load, p g represents the planned output of conventional units g, represents the abandoned air volume of wind turbine w in state s, f l max represents the upper limit of the transmission capacity of line l, represents the state of unit g under state s, ε UR , ε DR , ε L They represent the confidence level of not violating the upper reserve / lower reserve / line power flow constraints, G s represents the set of faulty units in state s, G b represents the set of conventional units on bus b, G represents the set of all conventional units, W represents the set of all wind turbines, and B represents the set of all buses.

[0108] Example:

[0109] In order to better understand the effect of the present invention, the proposed method for screening the expected system state is compared with the Nk criterion. The accuracy and computational efficiency of critical state screening are demonstrated by comparing the opportunity-constrained reliability-constrained economic dispatch model based on all expected system states, the Benders decomposition model based on all expected system states, and the random optimization model based on the scenario method of all expected system states. The accuracy of the proposed RCED model is demonstrated by comparing it with the adjustment constraint model without adding the replaced state.

[0110] The reliability-constrained economic dispatch method for power systems based on critical states proposed in this embodiment is tested on a modified IEEE RTS-79 system. RTS-79 consists of 32 generators, 38 transmission lines, with an installed capacity of 3405 MW and a peak load of 2850 MW. Generators 1, 6, 7, 8, 10, 11, 12, 13, 21, 25, 26, 27, 28, and 30 are replaced by wind farms of equal capacity. The goal is to use conventional and wind turbines in a cost-optimal manner to meet reliability requirements while meeting all capacity and flow constraints (with a high probability). In other words, we want to solve the problem. We consider LOLP to be 0.002, which means that the probability of no load loss in the system needs to exceed 0.998.

[0111] The cost coefficients, minimum and maximum generating capacities, and maximum reserve capacities of conventional generators are shown in Table 1. The reactance and capacity of the transmission lines are given in . The PTDF matrix Based on calculations. The goal of this embodiment is that the risk of insufficient system reserve capacity and line overload does not exceed 5% in any expected system state, so α is set UR ,α DR , and are all 0.05. The power forecast for each wind turbine is 70% of the rated capacity and the load forecast is 60% of the peak load. The random variable, i.e. the deviation between the forecast and the actual generation, is assumed to be a truncated normal distribution with mean zero and a truncation interval of 15% of the forecast value. The dispatch decision is assumed to be made about 15 minutes before the actual time when the demand is met. This justifies the limited support assumption for the deviation between the forecast and the actual realization. Note that technically, the wind turbines are able to produce from 0 MW to the rated capacity, which is considered to be their physical limit. All case studies were conducted on a PC equipped with an AMD Ryzen 74800H 2.9 GHz processor and 16 GB RAM. RCED utilizes the solver GUROBI’s YALMIP.

[0112] Table 1 Conventional power generation unit data

[0113]

[0114] At this point, the proposed method is compared with three other methods for solving the reliability constrained economic dispatch problem. These methods are explained as follows:

[0115] Improvement of reliability and solvability of scheduling model. First, compared with the Nk criterion, the improvement of reliability and solvability of scheduling model by the expected system state screening method is analyzed. In the Nk criterion, k = 1, 2, 3 is used.

[0116] The effectiveness of maximum probability state truncation. First, the speed of state truncation and the number of states were studied. By comparing the commonly used Nk criterion with the maximum probability state truncation method, the reduction of the number of states by the maximum probability state truncation was analyzed. In the Nk criterion, k = 1, 2, and 3 were used. The number and probability of states obtained by different state screening methods are given in Table 2. The Nk criterion does not consider the probability of the system state, and can quickly obtain the expected system state set. The expected system state generation time is 0.1498s, 0.1629s, and 0.2604s respectively. The proposed state truncation method takes 12.0094s to screen the state set that meets the probability requirements. However, the Nk criterion does not screen states according to probability, and will ignore high-probability states among high-order states. The total number of system states considered by the N-3 criterion is 29317, and the total state probability is 0.9892. The maximum probability state truncation method can take into account states with a higher probability of occurrence. The number of system states considered is 18848, and the total state probability is 0.999. Under the same LOLP requirement, the maximum probability state truncation method considers fewer states, reduces the states used for reliability constraints, and reduces model complexity. And unlike the discontinuity of the expected system set probability generated by the Nk criterion, the maximum probability state truncation method can filter the system state according to the LOLP requirement. Therefore, the use of the maximum state probability truncation method can effectively reduce the number of system states used for reliability constraints.

[0117] Table 2 The number and probability of states obtained by different state screening methods

[0118]

[0119] The necessity of removing the load loss state. The dispatch results obtained by different state screening methods are shown in Table 3. The N-1 criterion ensures that the system does not lose load with a probability of 0.7648, the unit's power generation cost is 112.7945, and the standby cost is 34.4406, but the reliability constraint considers fewer states and does not meet the LOLP requirements of system operation. N-2, N-3 criteria and truncation state The system state contains a system state that will still lose load no matter how it is scheduled. Therefore, the scheduling model has no solution. Although the proposed expected system state screening method takes 132.8055s to remove the state of losing load, the solvability of the model is guaranteed after removing the state of losing load, and the scheduling result that meets the system LOLP requirements can be obtained.

[0120] Table 3 Scheduling results obtained by different state screening methods

[0121]

[0122] Effectiveness and efficiency of critical state screening. As can be seen from Table 2, it takes 268.4032 seconds to screen the critical system state, which increases the time for state preprocessing, but reduces the number of states for establishing reliability constraints from 19,701 to 892. The solution time of Chance-and Reliability-Constrained Economic Dispatch (CRCED) based on the critical system state set is 32.3042 seconds, and it takes 300.7074 seconds from state screening to obtaining the scheduling result. The solution time of CRCED based on the expected system state is 728.5468 seconds, and it takes 861.3523 seconds from state screening to obtaining the scheduling result.

[0123] When establishing reliability constraints, each system state requires corresponding decision variables to establish corresponding upper and lower backup and line overload risk constraints. Therefore, the decision variables of the reliability constraints of the load loss state CRCED-AS are 20.72 times that of the critical system state CRCED-CS. Although the corresponding adjustment constraints are added to CRCED-CS to ensure that the represented state contains sufficient backup without load loss, it only increases the number of constraints without increasing the number of decision variables, which greatly reduces the modeling complexity and computational complexity.

[0124] Reliability and economy of critical system state screening. As can be seen from Table 3, the generation cost of removing the load loss state CRCED-AS and the critical system state CRCED-CS is 133.506, and the backup cost is 199.5185, which ensures that the system state set is under the risk of no more than 5% of the upper and lower backup and line overload risks in each state. Under the two methods, the unit output and standby have the same configuration, which proves that the proposed state screening method does not reduce the economy and reliability of the scheduling scheme while reducing the system state and speeding up the solution.

[0125] Necessity of critical system state screening. When the reliability level is further improved, such as the system LOLP requirement is 0.001. That is, the system is required to have a load-free probability greater than 0.999. The comparison of the results of removing the load-free state CRCED-AS and the critical system state CRCED-CS is shown in Table 4. After removing the load-free state CRCED-AS, the number of states used to establish reliability constraints increases to 36001, which increases the model solution time to 2328.7654s. In comparison, although the critical system state CRCED-CS takes 230.6757s more to screen the state, the number of states to establish reliability constraints is reduced to 1178, and the model solution time is also reduced to 35.7859s accordingly. Therefore, in large-scale systems, with the increase of LOLP requirements, the number of states that need to be considered increases dramatically. If the expected system state is not further screened, the complexity of the solution of the CRCED model will increase dramatically, which cannot meet the real-time requirements of system operation. Therefore, critical system state screening is very necessary to improve computational efficiency.

[0126] Table 4 Comparison of results between CRCED-AS and CRCED-CS

[0127]

[0128] In summary, the present invention reduces the number of states that need to be considered in the economic dispatch model by screening out the critical state set in the power system, thereby reducing the computational complexity of the model. While reducing the number of states, the high accuracy of the reliability index is ensured so that the reliability of the power system is not affected. The screened critical state set is used to construct the reliability constraints of the system. These constraints can ensure that no load loss occurs during the operation of the power system. Finally, the constructed reliability constraints are incorporated into the economic dispatch model, and the constraints of the model are modified or expanded so that the dispatch scheme in the dispatch model can achieve economic dispatch while satisfying the reliability.

[0129] The present invention screens system states according to LOLP index requirements, uses the opportunity-constrained DC optimal power flow model to remove states that will definitely lose load, then screens worse states in the state set according to the capacity of the faulty component, and finally, within the source-load prediction error range, screens out worse states in the set according to the tolerance of the system state to the output plan, thereby obtaining a critical state set. Through the above screening method, the number of states is effectively reduced, and the efficiency and accuracy of reliability-constrained economic dispatch of the power system are significantly improved.

[0130] The reliability-constrained economic dispatch method for power systems based on critical states designed by the present invention, which is applied to power system dispatching, constructs reliability constraints based on critical state sets, realizes accurate modeling of power system operation reliability, quickly and accurately evaluates power system operation reliability, and can be applied to reliability-constrained economic dispatching to improve system operation reliability. It can play a good supplementary role in the economic dispatching of traditional power systems and is of great significance to planning and dispatching that take reliability into account.

[0131] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described with reference to the preferred embodiments of the present invention, it should be understood by those skilled in the art that various changes may be made in form and details without departing from the spirit and scope of the present invention as defined in the appended claims.

Claims

1. A reliability-constrained economic dispatch method for power systems based on criticality, characterized in that: The steps include: S1. Obtain the LOLP index requirement of the power system, and obtain the cumulative probability requirement of the truncation state according to the obtained LOLP index requirement of the power system; S2. According to the cumulative probability requirement of the truncated state, the system state is selected based on the descending order of probability to obtain the system state set H; S3. According to the acquired system state set, a chance-constrained DC optimal power flow model is used to obtain a corresponding load loss amount to obtain a system state without load loss, thereby obtaining a state set I; S4, taking the upper and lower limits of each bus conventional unit of the system state in the calculated state set I and the upper and lower limits of the transmission capacity between each bus pair as the criterion of the good-bad relationship between the system states in the state set I, updating the state set I so that the system state in the state set I is the worst state, thereby obtaining the state set A; S5, dividing the system states in the state set A into different subsets according to whether the faulty units are the same, and screening each subset in turn to select the worst system state, thereby obtaining the critical state set C; S6. Based on the critical state set C, a reliability-constrained economic dispatch model with reliability constraints is constructed and solved to obtain a dispatch plan that meets the LOLP requirements.

2. The power system reliability-constrained economic dispatch method based on criticality according to claim 1, characterized in that: In step S1, the cumulative probability of the truncated state is required to be calculated by the following formula: In the formula, represents the cumulative probability requirement of the truncated state, represents the LOLP index requirement of the power system, and ΔP represents the slack variable that makes the screening state meet the LOLP index requirement of the power system.

3. The power system reliability-constrained economic dispatch method based on criticality according to claim 1, characterized in that: Step S2 includes: S21, initialize the cumulative probability P = 0, the currently selected system state set H is an empty set, and the currently selected maximum state order k = 0; S22, calculating the occurrence probability of each system state in the system state set, and arranging the occurrence probability of each system state in descending order Where S i,m represents the mth selected system state in the i-th order state, The probability of occurrence of the mth selected system state in the i-th order state; S23, according to the sorted system state list, select the system state ranked first in the system state list and add it to the system state set H, and update the cumulative probability of the current selection S24, when it is determined that the cumulative probability of the currently selected system state is greater than the cumulative probability of the truncated state, executing step S25; otherwise, executing step S26; S25, the system state selection stops, and the neighborhood state of the selected system state is added to the system state set H, and the final system state H is output; S26. Take the next state in the system state list as the currently selected system state, and return to step S23.

4. The power system reliability-constrained economic dispatch method based on criticality according to claim 3, characterized in that: In step S22, the calculation formula for calculating the occurrence probability of each system state in the system state set is as follows: Where P(s) represents the probability of a single system state s, X represents the set of all components in the power system, and X d represents the set of faulty components in the system state, p i represents the failure probability of component i.

5. The power system reliability-constrained economic dispatch method based on criticality according to claim 1, characterized in that: In step S3, the chance-constrained DC optimal power flow model is used to obtain the corresponding calculation formula for the load loss: In the formula, p g , They represent the minimum output, planned output and maximum output of the conventional unit g, respectively, and p w 、p w,f , They represent the planned output, predicted power generation and actual power generation of wind turbine w respectively; d b d b,f , They represent the planned supply value, predicted value, and actual value of the load on bus b, respectively. a represents the total deviation of injected power caused by source charge prediction error, They represent the power distribution factors of the conventional unit injection power, renewable energy injection power, and load demand converted into line power flow, respectively. represents the abandoned air volume of the wind turbine w, f l max represents the upper limit of the transmission capacity of line l, ε UR , ε DR , ε L They represent the confidence level of not violating the upper reserve / lower reserve / line power flow constraints respectively, G represents the set of all conventional units, W represents the set of all wind turbines, and B represents the set of all buses.

6. The power system reliability-constrained economic dispatch method based on criticality according to claim 1, characterized in that: In step S4, the calculation formula for the upper and lower limits of each bus conventional unit in the system state in the state set I is as follows: In the formula, represents the output upper limit of the conventional units on bus b under system state s, Indicates the lower limit of the output of the conventional unit on bus b under system state s, Indicates the state of conventional unit g under system state s; The calculation formula for the upper and lower limits of the transmission capacity between the bus pairs is as follows: In the formula, represents the upper limit of the transmission capacity between the bus pairs (k, b) in the system state s, represents the lower limit of the transmission capacity between the bus pairs (k, b) in the system state s, represents the state of line l in state s, and (k, b) represents a bus pair.

7. The power system reliability-constrained economic dispatch method based on criticality according to claim 1, characterized in that: Step S5 includes: S51, initializing the critical state set C to be an empty set; S52. Setting different subsets N g represents the number of types of faulty units, and the subset The first state in Set to the baseline state and initialize the data used to store the data from the sub-collection The state set filtered out is an empty set; S53, calculating the reference state Under the condition of no loss of load, the collection The maximum optimal load shedding LS c and the corresponding state c; S54, when LS c When LS is equal to 0, step S55 is executed; c If it is greater than 0, update for renew for Return to step S53; S55, will Updated to As the new baseline state, until the subset When each state in is taken as the reference state, the critical state set C is updated as 8. The power system reliability-constrained economic dispatch method based on criticality according to claim 7, characterized in that: In step S53, the reference state is calculated Under the condition of no loss of load, the collection The maximum optimal load shedding LS c And the corresponding state c is calculated as follows: maxLS c In the formula, p w,f They represent the actual value and predicted value of wind turbine w power generation, d b,f They represent the actual value and predicted value of busbar b load respectively. They represent the planned output of conventional unit g and wind turbine unit w under state s, respectively, and f l s represents the power flow of line l in state s, The state represents the phase angle of bus bar b in state s, represents the maximum value of the prediction error of wind turbine w, Indicates the maximum value of the load prediction error of bus b.

9. The power system reliability-constrained economic dispatch method based on criticality according to claim 1, characterized in that: In step S6, the reliability constraint is to impose a load-free constraint on each state in the critical state set C, and the specific steps are as follows: In the formula, represents the output fluctuation of conventional unit g caused by component failure in state s, represents the total deviation of injected power caused by source load prediction error in state s, represents the participation factor of the radiation control of the conventional unit g under state s, They represent upper standby and lower standby of conventional units respectively. They respectively represent the power flow distribution factors converted from conventional unit injection power / renewable energy injection power / load demand into line power flow under state s. p w They represent the actual power generation value and planned output of wind turbine w, Indicates the actual value of busbar b load, p g represents the planned output of conventional units g, represents the abandoned air volume of wind turbine w in state s, f l max represents the upper limit of the transmission capacity of line l, represents the state of unit g under state s, ε UR , ε DR , ε L They represent the confidence level of not violating the upper reserve / lower reserve / line power flow constraints, G s represents the set of faulty units in state s, G b represents the set of conventional units on bus b, G represents the set of all conventional units, W represents the set of all wind turbines, and B represents the set of all buses.

Images