Load emergency control method based on dynamic event triggering mechanism

Abstract

The present invention belongs to the field of smart power grids and power system automation control. Disclosed is a load emergency control method based on a dynamic event triggering mechanism, which method comprises: on the basis of an indicator of safe operation of a system, performing region division on a distribution network; proposing an adaptive load level division method based on user response willingness, and by taking the minimization of the system operation cost of the distribution network as an objective, establishing a precise load control model of the distribution network; and using a dynamic event triggering mechanism to reduce communication between substations and a master station of the distribution network, so as to realize effective exchange and precise execution of load control policy information. The method of the present invention can precisely realize precise control of a load, and can effectively alleviate communication transmission pressure of a distribution network, thereby ensuring the stability of a power grid and the continuous supply of critical loads.

Description

A load emergency control method based on a dynamic event triggering mechanism TECHNICAL FIELD

[0001] The present application belongs to the field of smart grid and power system automatic control, and particularly relates to a load emergency control method based on a dynamic event triggering mechanism. BACKGROUND

[0002] The rapid growth of power demand poses a severe requirement on power supply. When there is a power shortage in the distribution network line, the main station needs to make a decision to cut off part of the load quickly to stabilize the steady-state power shortage of the power grid and improve the safety resilience and regulation flexibility of the power grid. The traditional low-frequency load shedding control method cuts off the entire transmission line in the distribution network area to achieve the balance of power supply and use, which will increase the operation risk of the distribution network and have a negative impact on the normal economic development of society and the living standards of users. Therefore, it is necessary to study the load shedding in low-accident-level scenarios, and how to excavate the potential of the cuttable load on the distribution network side and accurately and quickly implement the load shedding operation is a difficult problem to be solved at present.

[0003] When the current distribution network performs accurate load shedding, the flexible resources on the user side and the response degree of the users under the guidance of time-of-use electricity price are not fully utilized. In addition, although the static event triggering mechanism can alleviate the communication pressure of the distribution network to a certain extent, the dynamic triggering mechanism is proved to be able to realize the transmission of information according to less data sampling or data transmission in a complex transmission system, thereby reducing the communication pressure. However, for the load shedding operation of the distribution network, how to set a suitable dynamic event triggering mechanism to reduce the communication pressure between the main station and the substation of the distribution network still needs to be studied. SUMMARY

[0004] To solve the above technical problems, the present application provides a load emergency control method based on a dynamic event triggering mechanism. In an emergency, the demand side, especially the willingness of livelihood load and the effective communication conditions, are fully considered. Through grid risk assessment and load priority division, the grid safety and stability control system realizes accurate cutting of the grid load according to the issued load shedding strategy, so as to ensure the stability of the grid and the supply of key loads.

[0005] The load emergency control method based on the dynamic event triggering mechanism comprises the following steps:

[0006] Step 1. According to the system safety operation index, the distribution network is divided into regions to ensure the conditions for the safe operation of the system;

[0007] Step 2. Based on the divided regions, an adaptive load shedding adjustment range method based on real-time electricity price is proposed, and a user participation load shedding response coefficient based on real-time electricity price is established;

[0008] Step 3, taking the minimum operation cost of the power distribution network substation as the target, a precise load control model of each region of the power distribution network is established;

[0009] Step 4, a dynamic event triggering mechanism is adopted to realize effective interaction and precise execution of the load shedding control strategy information of the region and its adjacent regions;

[0010] Step 5, according to the price signal, fully considering the willingness of the load shedding user side in the region to participate in load shedding, based on the precise load control model of the power distribution network, taking the minimum operation cost of the power distribution network as the target, taking the safe operation of the system as the constraint condition, using the alternating multiplier method to calculate the optimal load shedding strategy of each region;

[0011] Step 6, based on the optimal load shedding strategy obtained in step 5, when the power shortage occurs in the power distribution network, the interruptible load participating in the load shedding operation in the region is used to adjust the safe and stable operation of the system.

[0012] Further, in step 1, the power distribution network is divided into regions based on the consideration of the stability characteristics of the power grid, that is, the voltage regulation capability is fully considered on the basis of considering the modularity index, and the optimal partition of the power distribution network is realized;

[0013] The modularity index of the power distribution network is used to measure the correlation degree of two nodes in the power distribution network, which is as follows:

[0014]

[0015] In the formula, ρ l represents the correlation degree between nodes, l ij represents the electrical distance between node i and node j, i and j are nodes in the power distribution network, k i represents the node degree of node i, k i =∑ i l ij is the sum of the weights of all edges connected to node i; k j represents the node degree of node j; m is the sum of the weights of all edges in the region after partition, that is, m = (∑i∑ilij) / 2; δ(i,j) is a binary variable, δ(i,j) = 1 when node i and j are in the same region, otherwise δ(i,j) = 0;

[0016] The correlation degree index of two nodes in the power distribution network is as follows:

[0017]

[0018]

[0019] In the formula, i and j are both in the range of 1 to w, i≠j; S VP,jjSij represents the voltage sensitivity of the load node j to the active power source j VQ,jj Sij represents the voltage sensitivity of the load node j to the active power source j VP,ij Sij represents the voltage sensitivity of the load node j to the active power source j VQ,ij Sij represents the voltage sensitivity of the load node j to the active power source j iw Dwi represents the influence degree of the node w to the node i jw Dwj represents the influence degree of the node w to the node j ij Dij represents the comprehensive influence degree of the power change of the node j to the node i ij The smaller the value is, the greater the influence of the power change of the node j to the node i is, and the smaller the distance between the two nodes is l The greater the value is, the better the structure of the power distribution network division is

[0020] Considering the voltage regulation capability, a voltage regulation index is introduced for the regional division, and the specific process is as follows:

[0021]

[0022] In the formula, ρ V表 represents the average value of the voltage deviation in each region represents the voltage measurement index of each node i in the region n at time t; when the maximum voltage deviation ΔV i in the region is less than the maximum allowed voltage deviation ΔV i,max , otherwise, the value is ΔV i / ΔV i,max , and N represents the number of divided regions.

[0023] Further, in step 2, it is assumed that the expected value of the user at time t is p ex,t When the actual electricity price is higher than the user's expectation, the user's participation will be high, and then the adjustable amount of the load will increase; otherwise, the user's willingness will decrease, and the corresponding cuttable load will decrease, so the user response coefficient is established:

[0024]

[0025] In the formula, ρ i,t represents the response coefficient of the user located at the node i to the power distribution network at time t, p t represents the actual electricity price at time t, p max represents the maximum value of the electricity price, ξ i is the willingness of the user at the node i, ξ i ≥ 0

[0026] When ρ i,tWhen the value is greater than 0, users at node i actively participate at time t, increasing the upward adjustment margin of the load shelving at that node and consequently increasing the range of interruptible power adjustment. Conversely, when the value is less than 0, the downward adjustment margin of the load shelving will decrease, and consequently, the range of interruptible power adjustment will decrease. Therefore, a value is defined to reflect changes in the load shelving. for:

[0027]

[0028] when At that time, users actively participated, and At this time, the electricity price is lower than users' expectations, and user participation decreases accordingly; therefore, the upper and lower limits of the load shedding are adaptively adjusted based on the electricity price signal, i.e.:

[0029] when hour:

[0030]

[0031] when hour:

[0032]

[0033] The adjusted load shedding limit shall not exceed the total load value of the current users;

[0034] In the formula, This indicates the initial upper and lower limits of the load that can be sheared. This indicates the upper and lower limits of the load that users can cut off under the guidance of electricity prices.

[0035] Furthermore, the cost modeling for distribution network payments to users is as follows:

[0036]

[0037] In the formula, Let λ be the cost incurred by the distribution network in compensating users at time t, M be the number of interruptible load levels, and λ be the value of the load. m This represents the compensation price corresponding to the m-th level of interruptible load; the higher the load level, the greater the compensation the user receives, i.e., the compensation price λ corresponding to a higher load level. m The larger the value; Let m be the total interruptible load at time t.

[0038] The safe range constraint for interruptible load shedding is:

[0039]

[0040] In the formula The total amount of all interruptible loads located at node i at time t;

[0041]

[0042] In the formula The mth level interruptible load amount located at node i at time t;

[0043] The maximum cutting frequency constraint is:

[0044]

[0045] In the formula The maximum interruption frequency of the interruptible load is represented;

[0046] The minimum and maximum duration constraints of the cuttable load are:

[0047]

[0048] In the formula, b i,t The state of the interruptible load at time t is represented by b i,t-1 The state of the interruptible load at time t-1 is represented by b i,t When b The maximum interruption frequency of the interruptible load is represented; The maximum and minimum interruption duration of the interruptible load is represented.

[0049] Further, step 3 is specifically:

[0050] Step 3-1, establish a target function for the load shedding control of the distribution network:

[0051]

[0052] In the formula, The power value traded by node i with the main grid at time t is represented, which is positive when purchasing electricity from the main grid and negative otherwise; The cost coefficient of node i traded with the main grid at time t is represented;

[0053] To ensure the safe operation of the system, safety operation constraints need to be established, among which the line flow constraint is:

[0054]

[0055] In the formula The renewable energy output at node i at time t, the active power of energy storage charging and discharging, respectively, The corresponding reactive power is represented;

[0056] The phase angle relaxation of the branch power flow of formula (15) is carried out, and the linear power flow constraint obtained after the relaxation is as follows:

[0057]

[0058] The expression after the relaxation of the power at the head of the branch line is as follows:

[0059]

[0060] Formula (17) is further rewritten in the form of a rotating second-order cone programming as follows:

[0061]

[0062] To ensure that the controllable unit in the distribution network operates within the safe range of its active power, the upper and lower limits of its operation are constrained as formula (19):

[0063]

[0064] Similarly, to ensure that the controllable unit in the distribution network operates within the safe range of its reactive power, the upper and lower limits of its operation are constrained as formula (20):

[0065]

[0066] In the operation process of the distribution network, the voltage of each node needs to operate within its safe range, so the constraint condition of each node is formula (21):

[0067]

[0068] Further, step 4 is specifically:

[0069] Suppose the operation state value of the distribution network measured by the i-th measuring device at the k-th sampling time is x i,k =[P load,t ,P ij,t ,v i,t ] T The event generation function is expressed as:

[0070] l i (x i,k ,x i,τ )=||x i,k -x i,τ ||-σ i ||x i,τ || (22),

[0071] In the formula, x i,τ represents the system operation state vector transmitted to the distribution network master station at the latest time; and σ ia trigger threshold of the ith measurement device;

[0072] Based on formula (22), the generating function based on the dynamic trigger event mechanism is constructed as:

[0073] f(x i,k ,x i,τ )=αl(x i,k ,x i,k )-δ i,k (23),

[0074] In formula (23), δ i,k is a dynamic variable set in the dynamic trigger event mechanism, and the minimum sampling interval under the dynamic event trigger mechanism is not less than the minimum sampling interval under the static event trigger mechanism; δi ,k+1 =λδ i,k -l i (x i,k ,x i,τ ), λ is a scalar between (0, 1), and α≥1 / λ;

[0075] Based on formula (23), between the interval [t i,k ,t i,k+1 ] of any two trigger times, the logical trigger variable γ i,k is defined according to the calculation value of the event generating function, and the specific expression is:

[0076]

[0077] As can be seen from formula (24), when γ i,k is 1, the substation updates its own controller, and the master station transmits the state data of the substation to the adjacent distribution substation, l i (x i,k ,x i,τ ) will be set to 0, the trigger function f(·) <0, at this time the logical variable γ i,k is set to 0, and will be executed again when the trigger function again meets the next time, and the dynamic event trigger mechanism is realized when the power shortage occurs and other substation coordination is needed to cut the load to ensure stable operation of the system.

[0078] The method has the beneficial effects that the method guides users to actively participate in the load shedding operation of the distribution network by establishing a response coefficient under the current dynamic price mechanism, which not only helps to generate an adaptive load shedding strategy, but also enables users to obtain more compensation and reduces the operation cost of the distribution network master station. The dynamic event trigger mechanism is introduced to improve the information transmission efficiency between the distribution network master station and the substation, and effectively reduces the communication burden in the power information transmission process. BRIEF DESCRIPTION OF DRAWINGS

[0079] Fig. 1 is a flow chart of a dynamic event triggering mechanism;

[0080] Fig. 2 is a framework diagram of load shedding control strategy;

[0081] Fig. 3 is a schematic diagram of an improved 33-node distribution network system;

[0082] Fig. 4 is a graph of photovoltaic, wind turbine output power load demand and electricity price;

[0083] Fig. 5 is a schematic diagram of typical node voltage level;

[0084] Fig. 6 is a graph of all node voltage levels;

[0085] Fig. 7 is a schematic diagram of traditional proportional load shedding scheduling results;

[0086] Fig. 8 is a schematic diagram of adaptive load shedding scheduling results;

[0087] Fig. 9 is a schematic diagram of time triggering time of each sub-region;

[0088] Fig. 10 is a flow chart of the method of the present application. DETAILED DESCRIPTION

[0089] In order to make the content of the present application more easily and clearly understood, the present application will be further described in detail below according to specific embodiments and in conjunction with the accompanying drawings.

[0090] As shown in Fig. 10, the load emergency control method based on the dynamic event triggering mechanism comprises the following steps:

[0091] Step 1, according to the system safety operation index, the distribution network is divided into regions to ensure the conditions for safe operation of the system;

[0092] Step 2, based on the divided regions, an adaptive load shedding adjustment range method based on real-time electricity price is proposed, and a user participation load shedding response coefficient based on real-time electricity price is established;

[0093] Step 3, taking the minimum operation cost of the distribution network substation as the target, a precise load control model of the distribution network in each region is established;

[0094] Step 4, a dynamic event triggering mechanism is adopted to realize effective interaction and precise execution of the load shedding in the region and the load control strategy information of the adjacent region;

[0095] Step 5, according to the electricity price signal, fully considering the willingness of the load shedding users in the region to participate in load shedding, based on the precise load control model of the distribution network, taking the minimum operation cost of the distribution network as the target and taking the safe operation of the system as the constraint condition, an alternating multiplier method is used to calculate the optimal load shedding strategy of each region;

[0096] Step 6, based on the interruptible load optimization strategy obtained in step 5, when power shortage occurs in the distribution network, the interruptible load participating in the load shedding operation in the region is used to adjust the safe and stable operation of the system.

[0097] When the distribution network fails or power shortage occurs initially, problems such as frequency drop, power flow limit, and tie-line power overuse may occur, and the distribution network master control system should be able to realize fast, accurate and load operation. A typical load shedding control system is composed of a distribution network center master station, a distribution network substation and an electrical equipment control terminal, as shown in Figure 2. The distribution network decision layer master station is mainly responsible for receiving the information uploaded by the control center station, calculating the optimization decision for load shedding distribution, coordinating the data interaction between the modules of the distribution network substation, and issuing control operation instructions. The coordination layer of the distribution network substation mainly collects the interruptible load information of the region, interacts with the distribution network master station and the device layer, and executes the load shedding control instructions issued by the master station; the terminal device layer is provided with a control terminal to collect the interruptible load of the user end, upload it to the distribution network substation, and receive the instructions from the distribution network substation to realize fast load shedding operation.

[0098] When power shortage occurs in the power grid, the distribution network partition operation under low accident level can solve the problems of partial voltage and frequency limit, and cross-section power limit, etc. Compared with the overall load shedding measure, this method has the advantage of small control cost, so the distribution network is divided into regions when power shortage occurs on the basis of considering the stability characteristics of the power grid, that is, the voltage regulation capability is fully considered on the basis of considering the module index, and the optimized partition of the distribution network is realized.

[0099] The module index of the distribution network is used to measure the correlation degree of two nodes in the distribution network region, and is as follows:

[0100]

[0101] In the formula, l ij represents the electrical distance between node i and node j, i and j are nodes in the distribution network, k i represents the node degree of node i, k i =∑ i l ij is the sum of the weights of all edges connected to node i, and m is the sum of the weights of all edges in the region, that is, m=(∑i∑ilij) / 2; δ(i,j) is a binary variable, δ(i,j)=1 when node i and j are in the same region, otherwise δ(i,j)=0.

[0102] The correlation degree index of two nodes in the distribution network is as follows:

[0103]

[0104]

[0105] In the formula, S VP and S VQ respectively represent active and reactive voltage sensitivity matrix, D ij characterizes the degree of comprehensive influence of power change of node j on node i, D ij is smaller, the greater the influence of power change of node j on node i is, and the smaller the distance between the two nodes is; and the greater the value of ρ l is, the better the structure of the distribution network division is.

[0106] In addition to the physical meaning of the structure of the regional division of the distribution network, the stable operation of the system should also be concerned, therefore, the voltage regulation index is introduced as an important consideration factor of the regional division, the introduction of the index can effectively regulate the voltage out-of-limit problem when the power shortage occurs, and reduce the power loss between different regions:

[0107]

[0108] The index is represented by the average value of voltage deviation in each region, in the formula, V represents the voltage measurement index of each node i in region n, wherein when the maximum voltage deviation ΔV i in the region is less than the maximum allowed deviation ΔV i,max , the value is ΔV i / ΔV i,max , and N represents the number of divided regions.

[0109] Based on the time-of-use price, the user is guided to participate in the load shedding operation, and an adaptive load shedding strategy based on the user response coefficient is established. The user response coefficient is used to reflect the participation degree of the user, and the user benefit is used to reflect the compensation of the load shedding to the user.

[0110] Suppose that the expected value of the user at t time is p ex,t , when the actual price is higher than the expected value of the user, the participation degree of the user will be high, then the adjustable amount of the load will increase; on the contrary, the willingness of the user will decrease, and the corresponding cuttable load will decrease, therefore, the user response coefficient is established:

[0111]

[0112] In the formula, ρ i,t represents the response coefficient of the user at node i participating in the distribution network at t time, p t represents the actual price at t time, p max represents the highest value of the price, ξ i is the participation willingness of the user at node i, and ξ i ≥0; when ρi,t When > 0, the user at node i actively participates at time t, the up-regulation margin of the sheddable load at the node increases, and the corresponding interruptible power regulation range will increase accordingly; conversely, the down-regulation margin of the sheddable load will decrease, and the interruptible load regulation range will decrease accordingly. Therefore, the change value reflecting the sheddable load is defined as:

[0113]

[0114] From the foregoing analysis, it can be seen that, When the user actively participates, and When the electricity price is lower than the user's expectation, the user's participation degree decreases accordingly, and therefore the upper and lower limits of the sheddable load can be adaptively adjusted on the basis of the original values through the electricity price signal, that is, when When,

[0115]

[0116] When When,

[0117]

[0118] However, the modified upper limit of the sheddable load should not exceed the total load value of the current user; in the above formula, denote the initial upper and lower limits of the sheddable load power, denote the changed upper and lower limits of the sheddable load amount under the guidance of the electricity price, and the sheddable load amount is adaptively adjusted with the change of the electricity price, thereby increasing the flexibility of load shedding.

[0119] The present application realizes the load shedding function through the operation of the interruptible load. The user's participation in the load shedding is related to the interruptible load amount and the load level. The higher the level of the interruptible load, the higher the price of the compensation from the distribution network to the user, and from the perspective of the distribution network cost, the smaller the corresponding cost of the compensation is better. The cost paid by the distribution network to the user is modeled as follows:

[0120]

[0121] In the formula, M is the number of interruptible load levels, λ m is the compensation price corresponding to the mth level of interruptible load; in the method model, the higher the load level, the greater the compensation the user can obtain, that is, the higher the compensation price λ m of the high load level is. is the total amount of the mth level of interruptible load at time t.

[0122] In addition, the interruptible load should also meet the constraints of the sheddable load safety range, shedding times, shedding duration, etc.

[0123]

[0124]

[0125] The maximum resection times constraint is:

[0126]

[0127] The minimum and maximum duration constraints of the resection load are:

[0128]

[0129] In the formula, b i,t Indicates the state of the interruptible load, when b i,t = 1, indicates that the interruptible load participates in the load shedding operation, otherwise it does not participate; Indicates the maximum number of interruptions of the interruptible load; Indicates the maximum minimum interruption duration of the interruptible load.

[0130] The target function is established for load shedding control of the distribution network, and in addition to paying compensation fees to users for load shedding, the transaction cost with the main network is also considered for the distribution network master station:

[0131]

[0132] In the formula, Indicates the power value of node i traded with the main network at period t, which is positive when purchasing power from the main network, and vice versa; Is the cost coefficient of node i traded with the main network at period t.

[0133] In order to ensure the safe and stable operation of the distribution network, the Distflow optimal power flow is used to model the constraint condition, and the second-order cone programming method is used to relax the quadratic nonlinear condition in the Distflow model, and the original problem is converted into a mixed integer second-order cone programming problem.

[0134] Wherein the line flow constraint is:

[0135]

[0136] In the formula Indicate the renewable energy output, energy storage charging and discharging active power at i node at t time, Indicate the corresponding reactive power;

[0137] The phase angle of the branch flow of formula (15) is relaxed, and the linear flow constraint obtained after relaxation is as follows:

[0138]

[0139] The branch line head power relaxation expression is as follows:

[0140]

[0141] The formula (17) is further rewritten in the form of rotating second-order cone programming as follows:

[0142]

[0143] In order to ensure that the controllable unit active power in the distribution network runs in the safe range allowed, the upper and lower limits of its operation are constrained as formula (19):

[0144]

[0145] Similarly, in order to ensure that the controllable unit reactive power in the distribution network runs in the safe range allowed, the upper and lower limits of its operation are constrained as formula (20):

[0146]

[0147] In the operation process of the distribution network, the voltage of each node needs to run in the safe range, so the constraint condition of each node is formula (21):

[0148]

[0149] The objective function and constraint condition of the power load shedding model established by the application meet the form of second-order cone programming problem, the decision variable is the load switch state in each substation, and a commercial solver such as CLPEX can be used to effectively solve it.

[0150] When the power shortage occurs in the distribution network and the load shedding operation is performed, the system operation data collected by the measuring device needs to be transmitted to the distribution network substation through the communication network and the exact load shedding amount is quickly calculated, according to the willingness of users to participate in load shedding, the interruptible load participating in the load shedding operation in the region is used to adjust the safe and stable operation of the system.

[0151] When the region has not recovered to the system safe operation range, the demand information needs to be sent from the substation to the distribution network master station, and the distribution network master station selects the adjacent region which meets the conditions to carry out cross-region load shedding according to the load margin. In the implementation of load shedding operation, frequent transmission of information between the master station and the substation is needed, and the traditional mechanism usually adopts fixed period sampling. With the wide application of smart meters, the amount of measurement data transmitted in the network is increasing, and when transmission is carried out under the condition of limited bandwidth, the communication channel pressure will be increased, and the network-induced phenomenon will appear. In order to reasonably utilize the limited communication resources and relieve the network communication pressure, the dynamic event triggering mechanism is adopted, as shown in Figure 1, the event is defined as whether the adjacent sub-region participates in the load shedding operation, and the event triggering detector of the distribution network substation will act only when the trigger signal meeting the condition is received, then the event triggering controller will transmit the data information to the adjacent substation. The trigger function is set reasonably to reflect the state of the event being triggered, reduce the data transmission amount of the distribution master station in the implementation of load shedding operation, and relieve the communication pressure between the distribution network master station and the substation.

[0152] Suppose that the power distribution network operation state value measured by the ith measuring device at the kth sampling time is x i,k =[P load,t ,P ij,t ,v i,t ] T The event generation function can be expressed as:

[0153] l i (x i,k ,x i,τ )=||x i,k -x i,τ ||-σ i ||x i,τ || (22),

[0154] In the formula, x i,τ represents the system operation state vector transmitted to the distribution network master station at the latest time, since the dynamic triggering event is based on the static triggering event and adds an internal dynamic variable, the communication frequency between the system master station and the substation can be reduced, and the generation function based on the dynamic triggering event mechanism is constructed as:

[0155] f(x i,k ,x i,τ )=ɑl(x i,k ,x i,k )-δ i,k (23),

[0156] In the formula, δ i,kThe dynamic variable is set for the dynamic event triggering mechanism, and it is proved from the mathematical theory that the minimum sampling interval under the dynamic event triggering mechanism is not less than the minimum sampling interval under the static event triggering mechanism, so that unnecessary transmission can be further reduced in the information interaction process, where δ i,k+1 = λδ i,k - l i (x i,k , x i,τ ), λ is a scalar between (0, 1), and α ≥ 1 / λ.

[0157] In the interval [t i,k , t i,k+1 ] between any two triggering moments, the logical triggering variable γ i,k is defined according to the calculated value of the event generation function, and the value of γ i,k determines whether the state information of the substation is transmitted to the adjacent power distribution station through the communication network, and the specific expression of the measured event triggering logical variable is:

[0158]

[0159] It can be seen that when γ i,k is 1, the substation updates its own controller, and the master station transmits the state data of the substation to the adjacent power distribution station, l i (x i,k , x i,τ ) will be set to 0, the triggering function f(·) < 0, at this time the logical variable γ i,k is set to 0, and will be executed again when the triggering function satisfies again next time, and the dynamic event triggering mechanism is realized when the power shortage occurs and other substations are needed to coordinate load shedding to ensure stable operation of the system.

[0160] In order to verify the effectiveness of the load shedding method proposed in the application, the improved IEEE33 node power distribution network model is verified, as shown in Figure 3, wherein the node numbers are 1-33, the divided areas are ①②③, the reference capacity of the power distribution network system is 10MVA, the reference voltage is 12.66kV, and the voltage standard value of each node is allowed to be 0.95-1.05(pu). The information of wind turbine, photovoltaic, load and electricity price is shown in Figure 4.

[0161] In the simulation verification of the application, it is assumed that 11MW of power shortage occurs at node 12 during system operation, according to the partition method proposed in the application, the improved 33 node power distribution network is divided into three areas as shown by different colors in Figure 3.

[0162] The node 6, 14, 24 in the system accesses the first level interruptible load, 20, 23, 25-28 node accesses the second level interruptible load, 8-10 is the third type interruptible load. If only according to the importance of the load is removed, obviously, the load at node 6, 14, 24 is preferentially removed, which will cause the voltage of node 6 to exceed the upper limit, as shown in Figure 5, and the voltage of node 6 will exceed the upper limit at 11:00-13:00 and 15:00, which will cause part of the distributed power to trip off the grid. Since the method proposed in the application fully considers the voltage index of the region, and compensates different types of interruptible load according to the importance of the load, the load is dispersed in the process of power shortage, although part of the second and third level load is removed, but it can ensure that the voltage of the node where the distributed power is located does not exceed the limit, which can effectively avoid the risk of large-scale power failure.

[0163] The voltage level of the 33-node under the partition load shedding scheme proposed in the application is shown in Figure 6, all node voltages are within the safe range of 0.95-1.05 (p.u), which meets the voltage operation constraint of the distribution network after load shedding, compared with the traditional method according to the importance of the load, the partition load shedding proposed in the application can ensure the user power supply reliability of the distribution network in the process of load shedding operation.

[0164] In order to verify the correctness and effectiveness of the method proposed in the application, under the same test environment and parameter setting, the method proposed in the application is compared with the traditional method of load shedding operation according to the proportion, the results are shown in Figure 7 and Figure 8. At the same time, the interruptible load described in the application is divided into three levels, the higher the level, the more important the load. As can be seen from Figure 7 and Figure 8, when the system needs to cut the load of the user due to power shortage, the method used in the application obviously reduces the number of load shedding, and the method used in the application obviously reduces the number of important load shedding than the traditional method.

[0165] In the simulation verification, the expected price of the user is set to 467 yuan / MWh, which is between the maximum and minimum value of the grid price, which is reasonable. As can be seen from Figure 8, when the actual price is higher than the user's psychological price, the time period is (8:00-23:00), the user really participates in the load shedding operation mainly concentrates in (10:00-11:00, 14:00-18:00), as can be seen from Figure 8, the price of these time periods is much higher than the user's psychological expectation, so the user's enthusiasm for participating in load shedding is obviously higher than that of other time periods, which also shows that under the current time-of-use electricity price, the method proposed in the application can more effectively guide the user to actively participate in the distribution network load shedding operation. According to the different interruptible load range, the interruptible load can be automatically adjusted to realize the adaptive amount of interruptible load.

[0166] Because the user actively participates in the load shedding operation in the peak period of electricity price, and the higher the load level, the more compensation the user can obtain, compared with the traditional method, the method proposed in the application can enable the user to obtain more compensation, and the compensation amounts corresponding to primary load, secondary load and tertiary load are 350, 500 and 1000 (unit: yuan / MW) respectively. As can be seen from Fig. 8, the tertiary load amount released by the user at 11:00, 15:00 and 16:00 in the peak period of electricity price is 0.544 MWh, 0.595 MWh and 0.612 MWh, which is much larger than the important load released by the traditional method in these three periods. The tertiary load amount cut off by the traditional method in the peak period of electricity price is 0.16 MWh, 0.175 MWh and 0.18 MWh respectively, so the compensation obtained by the user for cutting off the load is relatively small. At the same time, based on the adaptive mechanism proposed in the application, the load that can be cut off changes dynamically according to the change of electricity price, which can make the operation cost of the distribution network smaller, because the cost of purchasing electricity from the grid is less than the compensation given to the user in the peak period of electricity price, which means that the distribution network master station only needs to pay less cost to purchase the released electricity from the user. Therefore, the method proposed in the application is more optimal under the same environment.

[0167] When the load shedding operation occurs due to the power shortage of the distribution network, the amount of load cut off in the region where node 12 is located cannot meet the demand, so signals need to be sent to the distribution network substations of the two adjacent regions, and finally the distribution network master station is uniformly coordinated to realize it. The measurement device in the region where node 12 is located transmits the current system operating state information (amount of load that can be cut off, branch power, node voltage value, etc.) as a trigger event to the distribution network substation and adjacent substation. Since the dynamic trigger event mechanism is adopted in the application to realize information transmission between adjacent regions, the data transmission rate can be significantly improved. Here, the method proposed in the application is compared with the static event trigger mechanism in the load shedding process. Under the same test environment and parameter settings, the simulation time of the two trigger mechanisms (dynamic event trigger and static event trigger) is 10s. The parameters λ and α are set to 0.6 and 1.5 respectively, and the trigger threshold σ is set to 0.01. The trigger times under the two mechanisms are shown in Table 1. It can be seen that the event trigger times are significantly reduced by using the method proposed in the application. Compared with the static trigger event, the trigger times are reduced by about 25% at the minimum and by about 46% at the maximum.

[0168] Table 1 Comparison of trigger times of different regions

[0169]

[0170] Fig. 9 shows the triggering time of the measuring device in the three regions under the dynamic triggering event triggering mechanism. In a given simulation time, due to the introduction of the dynamic disturbance δ, the interval between the two triggering times becomes larger, the maximum communication time interval of the adjacent sub-regions is 1.3s, 1.09s, 1.1s, the minimum communication time interval of each sub-region is greater than 0.05s, and it can be seen from Fig. 9 that the power shortage occurs in region ②, and the coordination of the load shedding of other two regions is needed to realize the stability of the system, the triggering times of region ② are obviously more than those of the other two regions, and it can be seen from Fig. 8 that the average communication time interval of the dynamic event triggering mechanism adopted in the application is longer, which will occupy less communication resources to some extent. It can be seen that the dynamic event triggering mechanism adopted in the application can greatly reduce the communication times of the load shedding information interaction between the distribution sub-regions, which is beneficial to reducing the data transmission pressure of the communication network to some extent, thereby improving the efficiency of the load shedding process.

[0171] The application is aimed at the deficiencies of the traditional load shedding scheme, and proposes a load emergency control strategy based on a dynamic event triggering mechanism. Under the current dynamic price mechanism, the response coefficient is established to guide users to actively participate in the load shedding operation of the distribution network, which not only helps to generate an adaptive load shedding strategy, but also enables users to obtain more compensation and reduces the operation cost of the distribution network master station. The application also improves the information transmission efficiency between the distribution network master station and the sub-station by introducing a dynamic event triggering mechanism, which effectively reduces the communication burden in the process of power information transmission.

[0172] The above only describes the preferred scheme of the application, and is not further limited to the application, and various equivalent changes made by using the content of the specification and drawings of the application are within the protection scope of the application.

Claims

1. A method for load emergency control based on a dynamic event triggering mechanism, characterized in that, Comprising the following steps: Step 1, according to the system safe operation index, the distribution network is divided into regions; Step 2, based on the partitioned region, an adaptive load shedding adjustment range method based on real-time electricity price is proposed, and a user participation load shedding response coefficient based on real-time electricity price is established; Specifically: Assume that the user's expected value of the electricity price at time t is p ex,t When the actual electricity price is higher than the user's expectation, the user's participation will be high, and the adjustable amount of the load will increase. Conversely, the user's willingness will decrease, and the corresponding cuttable load will decrease. Therefore, the user response coefficient is established: where ρ i,t denotes the response coefficient of the user located at node i participating in the power distribution network at time t, p t denotes the actual electricity price at time t, p max denotes the maximum value of the electricity price, ξ i is the participation willingness of the user at node i, ξ i ≥ 0. When ρ i,t > 0, the user at node i actively participates at time t, the up-regulation margin of the curable load at the node increases, and the corresponding interruptible power adjustment range will increase accordingly; otherwise, the down-regulation margin of the curable load will decrease, and the interruptible load adjustment range will decrease accordingly; therefore, the change value reflecting the curable load is defined as For: When At this time, the user actively participates, and At this time, the electricity price is lower than the user's expectation, and the user's participation decreases accordingly; Therefore, the upper and lower limits of the load that can be cut are adjusted adaptively based on the original basis through the electricity price signal, that is: When Time: When Time: The upper limit of the adjusted load that can be cut does not exceed the total load value of the current user; In the formulae, representing the initial upper and lower limits of the cuttable load power, Indicates the upper and lower limits of the change of the user's load that can be cut under the guidance of the electricity price; Step 3, taking the minimum operation cost of the distribution network substation as the target, a precise load control model of the distribution network in each region is established; Specifically: The objective function for load shedding control of power distribution network is established: In the formulae, Pit represents the power value traded by node i with the main grid at time period t, which is positive when the node buys power from the main grid and negative otherwise; a cost coefficient for node i to transact with the main grid at time period t; The cost of the distribution network to the user at time t, To ensure the safe operation of the system, the safe operation of the constraints need to be established, in which the line flow constraints are: In the formulae respectively represent the renewable energy output, the energy storage charging and discharging active power at the i-th node at time t, represents the corresponding reactive power; The total amount of all interruptible loads at node i at time t; The phase angle relaxation of the branch power flow is performed on the formula (15), and the linear power flow constraint obtained after the relaxation is as follows: The expression of the power relaxation at the head of the branch line is as follows: Equation (17) is further rewritten in the form of a rotated second-order cone program as follows: To ensure the controllable units in the distribution network operate within their allowed safe range of active power, the upper and lower limits of their operation are constrained as in equation (19): Similarly, to ensure that the controllable units in the distribution network operate within their allowed safe range of reactive power, the upper and lower limits of their operation are constrained as in equation (20): In the operation of the power distribution network, the voltage of each node needs to be operated within its safe range, so the constraint condition of each node is formula (21): Step 4, based on the dynamic event triggering mechanism, the effective interaction and precise execution of the load control strategy information of the region and its adjacent region are realized; Step 5, according to the electricity price signal, fully considering the willingness of the user side participating in load shedding in the region, based on the precise load control model of the distribution network, taking the minimum operation cost of the distribution network as the target, and taking the system safe operation as the constraint condition, the alternating multiplier method is used to calculate the optimal strategy of the load that can be cut in each region; Step 6, based on the optimal strategy of the load that can be cut obtained in step 5, when the power shortage occurs in the distribution network, the interruptible load participating in the load shedding operation in the region is used to adjust the safe and stable operation of the system.

2. The load emergency control method based on a dynamic event triggering mechanism according to claim 1, characterized in that, In step 1, the distribution network is divided into regions based on the consideration of the stability characteristics of the power grid, that is, the voltage regulation capability is fully considered based on the consideration of the modularity index, and the optimal partition of the distribution network is realized; The modularity index of the power distribution network is used to measure the correlation degree of two nodes in the distribution network region, and is specifically as follows: where ρ l represents the degree of association between nodes, l ij represents the electrical distance between node i and node j, i, j are nodes in the power distribution network, k i represents the node degree of node i, k i =∑l ij is the sum of the weights of all edges connected to node i; k j represents the node degree of node j; m is the sum of the weights of all edges in the region after partitioning, i.e. δ(i,j) is a binary variable, δ(i,j)=1 when node i,j is in the same region, otherwise δ(i,j)=0; The correlation degree index of two nodes in a power distribution network is as follows: where i and j are both from 1 to w, i≠j; S VP,jj represents the voltage sensitivity of the load node j to the active power source j, S VQ,jj represents the voltage sensitivity of the load node j to the reactive power source j, S VP,ij represents the voltage sensitivity of the load node i to the active power source j, S VQ,ij represents the voltage sensitivity of the load node i to the reactive power source j; D iw measures the influence of the node w on the node i, D jw measures the influence of the node w on the node j, D ij characterizes the comprehensive influence of the power change of the node j on the node i, D ij the smaller, the greater the influence of the power change of the node j on the node i, and the smaller the distance between the two nodes; p l the greater, the better the structure of the power distribution network division; Considering the voltage regulation capability, a voltage regulation index is introduced to divide the area, as follows: wherein p V represents the average value of the voltage deviation in each region; represents the voltage measurement index of each node i in region n at time t; when the maximum voltage deviation in the region ΔV i is less than the maximum allowed deviation of the region ΔV i,max , Otherwise the value is ΔV i / ΔV i,max N denotes the number of regions divided.

3. The load emergency control method based on a dynamic event triggering mechanism according to claim 1, characterized in that, The cost modeling for the network to pay to the user is as follows: In the formulae, to compensate the cost generated by the power distribution network to the user at time t, M is the number of interruptible load levels, λ m is the compensation price corresponding to the mth level of interruptible load; the higher the load level, the greater the compensation the user can obtain, that is, the compensation price λ m corresponding to the high load level is greater; The total amount of the mth interruptible load at time t; The interruptible load's cuttable load safety range constraint is: In the formulae the total amount of all interruptible loads located at node i at time t; In the formulae The amount of the mth interruptible load at node i at time t; The maximum number of resections constraint is: In the formulae Indicates the maximum interruption times of the interruptible load; The minimally resectable, maximum duration constraints are: wherein b i,t represents the state of interruptible load at time t, b i,t-1 represents the state of interruptible load at time t-1; when b i,t =1, it represents that the interruptible load participates in load shedding operation, otherwise it does not participate; represents the maximum number of interruptions of the interruptible load; Indicates the maximum and minimum interruption duration of the interruptible load.

4. The load emergency control method based on a dynamic event triggering mechanism according to claim 1, characterized in that, Step 4 is specifically: Assume that the i-th measurement device measures the operating state value of the power distribution network at the k-th sampling time as x i,k = [P load,t , P ij,t , v i,t ] T The event generation function is expressed as: l i (x i,k ,x i,τ )=||x i,k -x i,τ ||-σ i ||x i,τ ||(22), In the formula, x i,τ represents the system running state vector transmitted to the network master station at the latest time; σ i represents the trigger threshold of the i-th measuring device; Based on equation (22), the generating function based on the dynamic triggering event mechanism is constructed as: f(x i,k ,x i,τ ) = al(x i,k ,x i,k ) - δ i,k (23), In the formula, δ i,k is a dynamic variable set in the dynamic event triggering mechanism, and the minimum sampling interval under the dynamic event triggering mechanism is not less than the minimum sampling interval under the static event triggering mechanism; δ i,k+1 = λδ i,k -l i (x i,k ,x i,τ ), λ is a scalar between (0, 1), and ɑ≥1 / λ; Based on equation (23), the interval (t) between any two triggering times i,k ,t i,k+1 Between these points, a logical trigger variable γ is defined based on the calculated value of the event generation function. i,k The specific expression is: From equation (24), when γ i,k = 1, the substation updates its controller, and the main station transmits the state data of the substation to the neighboring substation, l i (x i,k , x i,τ ) will be set to 0, triggering the function f(·) < 0, at which time the logic variable γ i,k is set to 0, and the dynamic event triggering mechanism is implemented again when the power shortage occurs and other substation coordination is needed to cut the load to ensure system stability.

Images