CN-115882484-B - Electric automobile participation demand response control method and system

CN115882484BCN 115882484 BCN115882484 BCN 115882484BCN-115882484-B

Abstract

The invention relates to the technical field of load demand response control, in particular to a method and a system for controlling participation demand response of an electric automobile, wherein the method introduces the total power deviation delta P of system response load, improves a cost micro-increment rate model of a leading intelligent agent and a following intelligent agent, calculates the leading cost micro-increment rate eta i[k+1] of a k+1th iteration and the cost micro-increment rate eta i[k+1] of the following intelligent agent respectively according to the improved cost micro-increment rate model, calculates the response load index of each intelligent agent of the k+1th iteration by eta i[k+1] , calculates the total power deviation of response load of the leading intelligent agent, enters the next iteration loop if the deviation exceeds an allowable error range, and ends the iteration if the total power deviation delta P of response load does not exceed the allowable error range. The invention can quickly eliminate the total power deviation of the response load, and promote the peak clipping and valley filling effects, so that the dynamic control performance of the system is better and the response speed is faster.

Inventors

TONG ZEJUN
ZHANG CHUN
GAO PENGCHENG
WU SHUANG
LI HAOYU
DAI QIANCHENG

Assignees

安徽工程大学

Dates

Publication Date: 20260512
Application Date: 20221024

Claims (10)

1. The method for controlling the participation demand response of the electric automobile is characterized by comprising the following steps of: s101, starting; S102, calculating initial values of cost micro-increment rates ƞ i of a main intelligent agent and a plurality of following intelligent agents by initial loads, and determining a row random matrix Dn according to response load limits; s103, calculating the total power deviation delta P [ k ] of the real-time response load by the main intelligent agent; S104, judging whether the intelligent agent to be calculated is a dominant intelligent agent, if not, executing S107, and if so, executing S111; s105, distributing response load indexes according to the limit value; s106, the agent exits, and Dn is updated and recalculated; s107, calculating ƞ f [ k+1] of the following agent k+1 according to the formula (26); , (26) Wherein sigma is a convergence coefficient of following the intelligent agent, d ij is the (i, j) th element in the row random matrix Dn, and n represents the number of the intelligent agent; S108, calculating the response load quantity of the electric automobile cluster under the following agent; s109, judging whether the response load limit is exceeded, if yes, executing S105, and if not, executing S110; s110, judging whether all the following agent analysis calculations are completed, if not, executing S107, and if so, executing S114; S111 computing the first lead agent k+1st ƞ l [ k+1] according to equation (25) , (25) Wherein mu is a convergence coefficient of the main intelligent agent, d ij is the (i, j) th element in the row random matrix Dn, and n represents the number of the intelligent agents; s112, calculating the response load quantity of the electric automobile cluster under the main intelligent agent; S113, judging whether the response load limit is exceeded, if yes, executing S105, and if not, executing S114; s114, calculating response load indexes of all the agents, and accumulating; S115, calculating the k+1st real-time response load total power deviation delta P [ k+1 ]; S116, judging whether the delta P [ k+1] is within the allowable error range, if so, executing S117, and if not, executing S104; and S117, outputting response load indexes of each agent, and ending.
2. The method for controlling participation demand response of electric vehicles according to claim 1, wherein the electric vehicle cluster response load amount is obtained through calculation in the formula (23) in S108 and S112: , (23) Wherein P Di is the i-th electric automobile cluster response load quantity, and alpha i 、β i is the quadratic term and the first term coefficient of the demand side load response cost function.
3. The electric vehicle participation demand response control method according to claim 2, wherein the formula (23) is derived by substituting the formula (11) into the formula (17): , (11) (17) Defining the partial derivative eta i of C i to P Di as the cost incremental rate of the load response of the ith electric automobile cluster, C i as the load response cost of the ith electric automobile cluster and gamma i as the user benefit coefficient.
4. The method for controlling participation demand response of an electric vehicle according to claim 3, wherein the formula (11) is derived from a peak-clipping and valley-filling electric vehicle load response cost model, the peak-clipping and valley-filling electric vehicle load response cost model comprising: before electric vehicle load reduction is defined, the supply side income is: , , (1) Wherein p r is the retail electricity price at the demand side, the time-of-use electricity price mechanism causes p r to change over time, The basic load of the electric automobile is represented by n, which is the total number of electric automobile clusters; after the electric vehicle load is reduced, the cost of the demand response management of the rewards obtained by the user and equivalent to the demand side is expressed as: , , (2) Wherein, the 、 The quadratic term and the first term coefficients of the cost function are managed for the demand response respectively, For the benefit coefficient, if the user's enthusiasm is low and the load reduction demand is large, the value is large, The load quantity is reduced for the ith electric automobile cluster based on excitation type demand response at the moment T, wherein T is a time period set; After the load of the electric vehicle is reduced, the income of the supply side is as follows: , , (3) The demand response cost of the system for load shedding is as follows: , , (4) substituting the formulas (1) - (3) into the formula (4) to obtain the system peak clipping demand response cost as follows: , , (5) After the load of the electric automobile is increased, rewarding obtained by a user is demand response management cost of a demand side, and the rewarding is expressed as: , , (6) Wherein, the 、 The quadratic term and the first term coefficients of the cost function are managed for the demand response respectively, For the benefit factor, the lowest benefit equivalent to the user participating in the valley fill, Increasing the load quantity for the ith electric automobile cluster at the moment t; after the load of the electric automobile increases, the income of the supply side is: , , (7) the demand response cost of the system for increasing the load is as follows: , , (8) Substituting the formula (1), the formula (6) and the formula (7) into the formula (8) to obtain the system valley filling demand response cost as follows: , , (9) By combining the formula (5) and the formula (9), a load response cost model during peak clipping or valley filling of the system can be obtained, as shown in the formula (10): , , 。 (10)
5. the electric vehicle participation demand response control method according to claim 1, wherein the total load power deviation Δp is: (24) Wherein δP D is a total index of system response load, And responding to the cumulative value of the load quantity in real time.
6. The electric vehicle participation demand response control method according to claim 5, wherein the derivation process of the formula (25) and the formula (26) includes: Based on the traditional discrete average consistency protocol: (19) Wherein x i represents a consistent state variable of the ith agent, k is the iteration number, N i is a set of adjacent communication agents of the i agent, a ij is an element in an adjacent matrix A, a weight coefficient of a communication connecting line between the i agent and the j agent is represented, and when the communication connecting line exists, the weight coefficient is 1, otherwise, the weight coefficient is 0; element L ij in the laplace matrix L based on the adjacency matrix a: (20) The traditional discrete average consistency protocol expression is transformed into a discrete average consistency protocol expression shown in a formula (21) by using an element L ij in a matrix L according to the formula (19) and the formula (20): (21) (22) Wherein d ij is the (i, j) th element in the row random matrix Dn; formulas (25) and (26) are derived from formulas (21) and (24).
7. The method for controlling participation demand response of an electric vehicle according to claim 6, wherein the method comprises revising the cost micro-increment rate calculation formula of each agent according to the influence of the upper and lower limits of response load of the agent on the cost micro-increment rate limit value of the agent, as shown in formula (27); (27) Wherein eta u i and eta l i are respectively the upper limit value and the lower limit value of the cost micro-increment rate of the intelligent body i, epsilon is the convergence coefficient of the main intelligent body or follows the intelligent body, the upper limit and the lower limit of the response load of each intelligent body under different load responses are determined by the load capacity of the charging station and the number of signed user protocols, and the value is determined by an estimation and prediction method.
8. The method according to claim 6, wherein the step of determining whether the method falls into a dead cycle in S116 includes determining whether the method falls into a dead cycle, and if not, executing S104, and if so, issuing an alarm.
9. A system of an electric vehicle participation demand response control method according to any one of claims 1 to 8, comprising: The dispatching center layer collects the resource information of the responsable load quantity and dispatching cost of each intelligent agent on the premise of ensuring the dynamic balance of power and the stable voltage of nodes, so as to decide to send down peak clipping tasks in the peak period and valley filling tasks in the valley period of power utilization; The multi-agent cooperative control layer is composed of load agents, different electric automobile clusters are regulated and controlled by corresponding load agents, a leading load agent receives a load instruction and a total response load index which are cut down or increased by a dispatching center, electric automobile cluster resource information is collected and uploaded to the dispatching center layer, real-time response load quantity of each electric automobile cluster in the electric automobile response layer is collected and used for calculation, and the following load agents cooperate with the leading load agent to jointly complete a demand response cooperative task, so that peak clipping and valley filling are realized; The load intelligent agent transmits load reduction or increase control signals to the electric automobile response layer, and the electric automobile response layer is an electric automobile cluster formed by a plurality of electric automobiles in one area.
10. The system of claim 9, wherein the electric vehicle cluster adopts a centralized communication mode.

Description

Electric automobile participation demand response control method and system Technical Field The invention relates to the technical field of load demand response control, in particular to a method and a system for controlling participation demand response of an electric automobile. Background With the rapid development of electric automobile technology and charging facilities, and in order to achieve the aim of double carbon, the exploitation amount of fossil energy is reduced, and the number of electric automobiles in the future is increased drastically. The load charging operation of a large number of electric vehicles can cause multi-node load power disturbance, so that the voltage stability of the public coupling point is influenced, and the cooperative control of the hybrid micro-grid group is influenced. In order to solve the problems, part of the technology utilizes household electricity loads such as air conditioners and the like to participate in system demand response control based on a discrete master-slave consistency algorithm, and establishes a multi-agent energy management system. However, the algorithm cannot quickly eliminate the dynamic deviation of the total power, so that the problems of non-ideal peak clipping and valley filling, poor dynamic control performance of the system and slow response speed are caused. Disclosure of Invention Therefore, the invention aims to provide the electric automobile participation demand response control method, so as to solve the problems of unsatisfactory peak clipping and valley filling, poor system dynamic control performance and slow response rate caused by the fact that the traditional consistency algorithm cannot quickly eliminate the total power dynamic deviation. Based on the above object, the present invention provides a method for controlling participation demand response of an electric vehicle, comprising: s101, starting; s102, calculating initial values of cost micro-increment rates eta i of a main intelligent agent and a plurality of following intelligent agents by initial loads, and determining a row random matrix Dn according to response load limits; S103, calculating the total power deviation delta P [ k ] of the real-time response load by the main intelligent agent; S104, judging whether the intelligent agent to be calculated is a dominant intelligent agent, if not, executing S107, and if so, executing S111; s105, distributing response load indexes according to the limit value; s106, the agent exits, and Dn is updated and recalculated; S107, calculating the kth+1st-order eta f[k+1] of the following agent according to the formula (26); Wherein sigma is a convergence coefficient of following the intelligent agent, d ij is the (i, j) th element in the row random matrix Dn, and n represents the number of the intelligent agent; S108, calculating the response load quantity of the electric automobile cluster under the following agent; s109, judging whether the response load limit is exceeded, if yes, executing S105, and if not, executing S110; s110, judging whether all the following agent analysis calculations are completed, if not, executing S107, and if so, executing S114; S111 calculating the kth+1st η of the primary agent according to equation (25) l[k+1] Wherein mu is a convergence coefficient of the main intelligent agent, d ij is the (i, j) th element in the row random matrix Dn, and n represents the number of the intelligent agents; s112, calculating the response load quantity of the electric automobile cluster under the main intelligent agent; S113, judging whether the response load limit is exceeded, if yes, executing S105, and if not, executing S114; s114, calculating response load indexes of all the agents, and accumulating; s115, calculating the total power deviation delta P [k+1] of the k+1st real-time response load; S116, judging whether the delta P [k+1] is within the allowable error range, if so, executing S117, and if not, executing S104; and S117, outputting response load indexes of each agent, and ending. Optionally, in S108 and S112, the electric vehicle cluster response load amount is obtained through calculation in a formula (23): Wherein P Di is the i-th electric automobile cluster response load quantity, and alpha i、βi is the quadratic term and the first term coefficient of the demand side load response cost function. Optionally, the formula (23) is derived by substituting formula (11) into formula (17): Defining the partial derivative eta i of C i to P Di as the cost incremental rate of the load response of the ith electric automobile cluster, C i as the load response cost of the ith electric automobile cluster and gamma i as the user benefit coefficient. Optionally, the formula (11) is derived according to a peak-clipping and valley-filling electric vehicle load response cost model, and the peak-clipping and valley-filling electric vehicle load response cost model includes: before electric vehicle load reduction is def