Air conditioner load cluster control and scheduling optimization method

Abstract

The application discloses an air conditioner load cluster control and scheduling optimization method, comprising the following steps: constructing an air conditioner load state queue model according to an equivalent thermal parameter model of a single air conditioner; constructing a dynamic response model of the air conditioner load group state queue model according to the influence of temperature control on the aggregated power of the air conditioner; and completing scheduling optimization of the aggregated air conditioner load participating in the demand response process according to the power characteristics of the dynamic response model and the comfort requirement of users. The method starts from improving the control accuracy of the aggregated power of the air conditioner load, comprehensively considers the suppression of rebound load and the consistency of the comfort of controlled air conditioner users, and provides a scheduling optimization strategy allowing air conditioner users to participate in demand response for multiple times.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of air conditioning load control, and particularly relates to an air conditioning load cluster control and scheduling optimization method. BACKGROUND

[0002] With the large-scale grid connection of intermittent renewable energy (such as wind energy and solar energy) and new grid loads (such as electric vehicles), the contradiction between supply and demand of the power system is increasingly prominent, and the safe and stable operation of the power system faces great challenges. Under the traditional grid operation mode, it often needs to pay a high cost to adjust the output of the supply-side generator set, so it is of higher application value to guide users to participate in demand response and tap the adjustable potential of the load side to relieve the power supply pressure of the grid. Among various flexible loads that can interact with the grid in a bidirectional manner, the air conditioning load has the characteristics of thermal energy storage, flexible scheduling mode, large user scale, fast response speed, strong controllability and the like, and can provide a large adjustment potential without affecting or with little effect on the user experience. In recent years, the rapid development of the bidirectional communication technology and advanced measurement system of the smart grid provides technical support for monitoring and controlling the user-side load, and is also the equipment basis for implementing demand response operation.

[0003] Implementing a reasonable control strategy for a large amount of air conditioning loads can reduce the peak load pressure of the power system, ensure the safe and stable operation of the system, and has good social and economic benefits. The research on the participation of air conditioning loads in demand response mainly includes two aspects of control mode and scheduling optimization, the former focuses on the fine control method of aggregated air conditioners or single air conditioners, and strives to suppress load fluctuations and load rebound, and the latter concentrates on evaluating the response potential on the basis of air conditioning load aggregation modeling and selecting appropriate scheduling strategies to maximize social benefits. Direct load control (DLC) is a demand response control method based on incentives in the electricity market, and is also the main control mode for air conditioning loads to participate in demand response. When implementing the DLC project, the users can be remotely regulated and controlled through strategies such as switch control and temperature control. The switch control has fast response speed and large short-time regulation potential, but it is easy to cause large power fluctuations in the control process, while the temperature control can better guarantee the user experience and meet the regulation demand in a long time scale, so it is applied in a large number of researches.

[0004] The existing mature technologies include:

[0005] 1. The state queueing model (SQ model) is used to analyze the reason for the oscillation of the aggregated power of air conditioning loads, and a method of separating the upper and lower limits of the temperature set value is proposed to smooth the power fluctuations during the response;

[0006] 2. The mechanism of load diversity destruction is analyzed, and the temperature set value of each air conditioner is set independently to effectively smooth the load fluctuation by combining switch control and temperature control. However, the above literature does not consider the difference between air conditioner parameters and user comfort;

[0007] 3. The power drop phenomenon that occurs when the air conditioner load is collectively controlled is modeled, and a correction coefficient is proposed to reduce model error. However, the change of air conditioner load control period under different temperature set values is ignored, and the error in the aggregation model greatly affects the scheduling result in large-scale air conditioner load centralized control.

[0008] 4. The dynamic response process of the aggregated power at the initial stage of temperature control is comprehensively analyzed, but the changes of the aggregated power in the steady stage and temperature recovery stage are not discussed.

[0009] 5. Based on the first-order equivalent thermal parameter model, an approximate aggregation model of temperature-controlled load is established, the influence of air conditioner parameters, temperature set value and ambient temperature on the aggregated power is analyzed, and a response potential evaluation method considering the uncertainty of user participation response is proposed;

[0010] 6. The lock-in time constraint of air conditioner on-off state switching is introduced, which improves the accuracy of adjustable capacity evaluation;

[0011] 7. The control process is divided into response control and load recovery stages, and the scheduling optimization is carried out respectively with the highest user comfort and the shortest temperature recovery time as the target. However, the decision model needs to judge whether each air conditioner is controlled or not, which has a large amount of calculation, and the air conditioner cannot be adjusted back once it is controlled in the response period, resulting in a large difference in the length of time for different users to participate in the control;

[0012] 8. The optimal scheduling plan is made in the day-ahead to guide the selection of scheduling strategy for the next day aggregation group, and the preparation time is introduced to reduce the impact of large load drop at the initial stage of response. However, this scheduling strategy sacrifices the control potential of the aggregated power drop stage, and does not consider the temperature recovery process.

[0013] In summary, the existing research can be seen as:

[0014] (1) The existing research lacks classification discussion and complete modeling of the dynamic response of the aggregated power in each stage of the temperature control process, including the power drop stage, the power stable stage and the temperature recovery stage;

[0015] (2) The influence of air conditioner parameter difference and temperature set value change on air conditioner start-stop cycle is ignored, resulting in a large error in the established aggregated power model;

[0016] (3) At present, most of the scheduling strategies stipulate that only one air conditioner is allowed to participate in the regulation during the demand response, which limits the exertion of the air conditioning load regulation potential, and the comfort experience difference between users is large under the regulation mode. SUMMARY

[0017] In view of the above problems in the prior art, the air conditioning load cluster control and scheduling optimization method provided by the present application solves the above problems.

[0018] In order to achieve the above-mentioned purposes, the technical scheme adopted by the present application is as follows: an air conditioning load cluster control and scheduling optimization method, comprising:

[0019] According to the equivalent thermal parameter model of a single air conditioner, an air conditioning load state queue model is constructed;

[0020] According to the influence of temperature control on the aggregated power of air conditioners, a dynamic response model of the air conditioning load group state queue model is constructed;

[0021] According to the power characteristics of the dynamic response model and the user comfort demand, the scheduling optimization of the aggregated air conditioning load participating in the demand response process is completed.

[0022] Further, the air conditioning load state queue model divides the start-stop cycle of the air conditioner into t off +t on state units according to the time interval Δt;

[0023] The first t off state units represent that the air conditioner compressor is in the shutdown state, and the indoor temperature gradually rises, and the last t on state units represent that the air conditioner is in the start-up cooling state.

[0024] Every time interval Δt, the air conditioners in each state unit are uniformly transferred to the next state, and the number of air conditioners contained in each state group is N / (t off +t on) .

[0025] Wherein, N is the total number of air conditioners.

[0026] Further, the dynamic response model of the air conditioning load group state queue model divides the change process of the aggregated power of the air conditioning load after responding to the control into:

[0027] The load reduction process during the response: including the power drop stage and the power stable stage of the demand response;

[0028] The power drop stage of the demand response is used to represent the case that the number of air conditioner load start state groups and the aggregated power change with time from the t0 moment when the regulation instruction is issued to the t3 moment when the first state group of the state queue cools down to the new temperature lower limit;

[0029] The state group represents the air conditioner in the state unit;

[0030] The power smooth stage is used to represent the case that the number of air conditioner load start state groups and the aggregated power change with time when the air conditioner state group runs in the new temperature range [T min,2 ,T max,2 ], corresponding to the t3 to t recv moment; it includes multiple complete cycles;

[0031] Each complete cycle includes: the t3 moment when the first state group cools down to the new temperature lower limit to the t7 moment when all the vacancy / overlap state groups have all been switched to the shutdown state;

[0032] Temperature recovery process after the response ends: used to represent the process that the temperature of the air conditioner cluster cools down from the higher temperature range [T min,2 ,T max,2 ] to the lower temperature range [T min,1 ,T max,1 ].

[0033] Further: the power smooth stage is modeled according to the correlation of the time required for the first and last state groups of the state queue to reach T min,2 , which includes three cases:

[0034] Case 1: the time for the first state group of the state queue to reach T min,2 is equal to the time for the last state group of the state queue to reach T min,2 ;

[0035] Case 2: the time for the first state group of the state queue to reach T min,2 is greater than the time for the last state group of the state queue to reach T min,2 ;

[0036] Case 3: the time for the first state group of the state queue to reach T min,2 is less than the time for the last state group of the state queue to reach T min,2 ;

[0037] Wherein, T min,2 represents the minimum value of the indoor temperature after warming up.

[0038] Further: in the temperature recovery process after the response ends, the temperature set value of the air conditioner in the shutdown state is set to the temperature set value after warming up;

[0039] For the air conditioner in the start-up state, the regulation is carried out according to whether there is a state overlap phenomenon or not:

[0040] If there is no state overlap phenomenon, the temperature setting value is set as the temperature setting value before the temperature rise;

[0041] If there is a state overlap phenomenon, the overlapping state groups are separated and then the regulation is carried out respectively.

[0042] Further: the method for scheduling optimization of the dynamic response model of the load group state queue model comprises:

[0043] The air conditioner is clustered and divided into K aggregated groups, and the user comfort degree is quantified by predicting the average vote;

[0044] A scheduling optimization model for one-time response of the user is constructed according to the user comfort degree;

[0045] The scheduling optimization model for one-time response of the user is optimized to obtain a scheduling optimization model for multiple responses of the user.

[0046] Further: the scheduling optimization model for one-time response of the user comprises a first stage and a second stage;

[0047] The first stage schedules and optimizes the load reduction process during the response period, and the optimization target is to minimize the deviation between the actual load reduction amount and the demand amount;

[0048] The second stage schedules and optimizes the temperature recovery process after the response, and the optimization target is to keep the aggregated power of the temperature recovery process within the rebound load limit value and keep the power stable.

[0049] Further: the way for optimizing the scheduling optimization model for one-time response of the user comprises:

[0050] The air conditioner controlled by the temperature ends the response in advance;

[0051] The regulation times of each aggregated group are limited;

[0052] The number of air conditioners in the controlled state at any time is limited to not more than the total number of air conditioners in the aggregated group;

[0053] The priority of the air conditioner user participating in the response is sorted.

[0054] The beneficial effects of the present application are:

[0055] 1. The change of the aggregated power in the power drop stage, the power stable stage and the temperature recovery stage is completely described, especially when there is a state overlap / void phenomenon, and the temperature recall time and specific control process for maintaining temperature diversity are pointed out;

[0056] 2. Further grouping according to temperature setting initial value not only ensures that the aggregated power remains stable after temperature adjustment, but also reduces the load rebound caused by simultaneous temperature adjustment of the air conditioner to a certain extent;

[0057] 3. The dispatching strategy of selecting the air conditioner entering temperature control in batches in different time periods during the demand response period can avoid the power valley in the initial response period, and the mutual power fluctuation suppression among the aggregated groups in the power drop stage and the power stable stage can be comprehensively utilized to effectively reduce the load reduction deviation;

[0058] 4. The dispatching strategy combining the multiple rotation of air conditioner users in demand response not only reduces the influence of temperature control on the comfort of air conditioner users to a certain extent, but also reduces the comfort difference between users. BRIEF DESCRIPTION OF DRAWINGS

[0059] Figure 1 Flow chart of air conditioner load cluster control and dispatching optimization method.

[0060] Figure 2 Equivalent thermal parameter model diagram of a single air conditioner load.

[0061] Figure 3 Schematic diagram of the change of room temperature, on-off state and power of an air conditioner with time.

[0062] Figure 4 Schematic diagram of air conditioner load state queue model.

[0063] Figure 5 Schematic diagram of state transition process of air conditioner load.

[0064] Figure 6 Schematic diagram of load reduction process during response period.

[0065] Figure 7 Schematic diagram of three cases considered in the power stable stage.

[0066] Figure 8 Diagram of simulation verification result of aggregated power model in the embodiment.

[0067] Figure 9 Comparison diagram of air conditioner aggregated power with different temperature setting initial values when the temperature is increased by 1℃ in the embodiment.

[0068] Figure 10 Comparison diagram of aggregated power with or without grouping based on temperature setting initial value in the embodiment.

[0069] Figure 11 Result diagram of dispatching optimization scheme 4 and dispatching optimization scheme 5 in the embodiment.

[0070] Figure 12The scheduling effect diagram of scheduling optimization schemes 3, 4 and 5 under different demand reduction ratios in the examples.

[0071] Figure 13 The group aggregation power comparison diagram of schemes 3 and 4 under a 10% reduction ratio in the fifth aggregation group in the examples. DETAILED DESCRIPTION

[0072] The specific embodiments of the present application are described below to facilitate the understanding of the present application for those skilled in the art, but it should be clear that the present application is not limited to the scope of the specific embodiments, and for those skilled in the art, it is obvious that various changes are within the spirit and scope of the present application defined and determined by the appended claims, and all the inventions utilizing the concept of the present application are within the scope of protection.

[0073] As shown in Figure 1 , in one embodiment of the present application, a method for air conditioner load cluster control and scheduling optimization is provided, comprising:

[0074] According to the equivalent thermal parameter model of a single air conditioner, an air conditioner load state queue model is constructed;

[0075] According to the influence of temperature control on air conditioner aggregation power, a dynamic response model of the air conditioner load group state queue model is constructed;

[0076] According to the power characteristics of the dynamic response model and the user comfort demand, the scheduling optimization of the aggregated air conditioner load participating in the demand response process is completed.

[0077] As shown in Figure 2 , the equivalent thermal parameter model diagram of a single air conditioner load, from the perspective of simplified calculation, makes the indoor air temperature T a equal to the indoor solid temperature T m , and further obtains a first-order ETP model representing the indoor temperature change and the air conditioner switch state:

[0078]

[0079] Wherein, C is the equivalent gas specific heat capacity, R is the equivalent thermal resistance; Q = ηP rated is the refrigeration / heating capacity of the air conditioner, η is the air conditioner energy efficiency ratio, P rated is the rated power of the air conditioner, s(t) is the switch state of the air conditioner at time t;

[0080] The s(t) is developed for the refrigeration mode of the fixed frequency air conditioner, the air conditioner compressor is periodically started and stopped in a certain environment, and the corresponding indoor temperature periodically changes in a certain range. The basic principle is that when the indoor temperature is higher than the upper limit of temperature, the air conditioner compressor is started, and the indoor temperature gradually decreases; when the indoor temperature is lower than the lower limit of temperature, the air conditioner compressor is turned off, and the indoor temperature gradually rises, so s(t) is expressed as:

[0081]

[0082] In the formula: T max and T min respectively represent the upper and lower limit values of the temperature of the air conditioner, which are determined by the temperature setting value T set and the temperature dead zone δ.

[0083] Combining the above analysis, the discrete calculation formula for describing the change of indoor temperature in the periodic operation process of the fixed frequency air conditioner can be obtained:

[0084]

[0085] Where Δt represents the simulation time interval.

[0086] Further, considering that the air conditioner runs in the interval of [T min , T max ], and assuming that the outdoor temperature remains unchanged in the control period, according to formula (3) and formula (4), the on-off time of a single air conditioner load in a single start-stop cycle is obtained:

[0087]

[0088] Where τ off represents the off time of the air conditioner compressor, and τ on represents the on time of the air conditioner compressor.

[0089] As shown in Figure 3 , the instantaneous power of the fixed frequency air conditioner depends on the rated power P rated of the air conditioner and the on-off state of the compressor, the on-off state of a single air conditioner, the room temperature, and the change of the instantaneous power with time, the aggregated total power P agg (t) of the air conditioner load cluster with consistent or similar parameters is as shown in formula (7):

[0090]

[0091] Where N on (t) represents the number of air conditioners in the on state at time t.

[0092] When the number of air conditioners N is large enough and each air conditioner load is in a steady state (i.e., the temperature set value remains unchanged), it is considered that the air conditioner cluster has high load diversity, i.e., the indoor temperature of each user is uniformly distributed in the temperature interval [T min ,T max ]. The state queue model is introduced to describe the aggregated power variation process of the air conditioner load, as shown in Figure 4 . The start-stop cycle of the air conditioner is divided into t off +t on state units according to the time interval Δt; the first t off state units represent that the indoor temperature gradually rises when the compressor is in an off state, and the last t on state units represent that the air conditioner is in an on state for cooling; each state unit is uniformly transferred to the next state every time interval Δt, and the number of air conditioners in each state group is N / (t off +t on) ; wherein N is the total number of air conditioners.

[0093] In Figure 4 , t off is 7, so the first 7 state units represent off, t on is 5, so the last 5 state units represent on, and there are a total of 12 state units.

[0094] Therefore, the aggregated power variation of the air conditioner load is only determined by the number of state groups in the on state under the current parameter condition, and is independent of the load number, i.e.,

[0095]

[0096] wherein S off (t) and S on (t) represent the number of state groups in the off and on states at time t, respectively.

[0097] Direct load control (DLC) is an incentive-based demand response control method in the electricity market, and is also the main way for air conditioning load to participate in demand response. When implementing the DLC project, various control strategies such as switch control and temperature control can be used to remotely control the users. Switch control is a control method in which the relevant departments send a shutdown instruction to the air conditioning load in advance to control the on-off state of the air conditioner, and the air conditioner is turned on again after the response ends, so as to achieve the purpose of load reduction or transfer. This control method has large capacity and fast response, but it cannot reduce the load for a long time due to the limitation of user comfort, and a large number of controlled loads are put into operation at the same time after the demand response ends, resulting in a rapid and severe impact of rebound load, which will inevitably have an adverse effect on the operation of the power grid. The principle of load reduction of temperature control is to change the temperature setting value of the air conditioner to reduce the on-off ratio of the air conditioner in the periodic operation process. As can be seen from formulas (5) and (6), when the external environment temperature remains unchanged, the shutdown and startup time of the air conditioner compressor will change with the increase and decrease of the temperature setting value. According to formula (8), the aggregated power of the air conditioning cluster also changes accordingly. This control method has smaller load capacity but can meet the control requirements in a long time scale.

[0098] The traditional temperature regulation method will cause a huge power fluctuation of the aggregated power, because the command of changing the temperature setting value of the air conditioning load cluster with the same parameters uniformly will seriously destroy the continuity and diversity of the state queue.

[0099] Therefore, in order to keep the aggregated power relatively stable, the SP control method is adopted in the present application, which sets the temperature setting value for the air conditioner in the startup state and the shutdown state respectively, so that the state queue runs between the initial temperature lower limit and the temperature upper limit after warming up, ensuring the continuity and temperature diversity of the state transition process. The control method is realized by establishing a dynamic response model of the air conditioning load cluster state queue model;

[0100] Because the start-stop cycle (total number of states) and the startup and shutdown time (state distribution) of the air conditioner are different when it operates at different temperature setting values, the following data before and after the temperature of the air conditioning load is warmed up need to be calculated before the dynamic response model of the air conditioning load cluster state queue model is constructed under the temperature control method:

[0101]

[0102] Wherein, S off,m represents the number of shutdown and startup state units of the air conditioning load in one start-stop cycle when the temperature setting value is T set,m (m = 1, 2); T set,1 and T set,2 represent the temperature setting values before and after warming up, respectively; S on,m represents the corresponding indoor temperature range [T min,m,T max,m ] and T set,m Satisfy relation (3); ΔF 1,2 Indicates air conditioning from T max,1 Heat up to T max,2 Number of time intervals required, ΔN 2,1 Indicates air conditioning from T min,2 Cool down to T min,1 The required number of time intervals.

[0103] In this application, the dynamic response model of the air conditioning load group state queue model divides the aggregate power change process of the air conditioning load after response control into:

[0104] The load reduction process during the response period includes the power drop phase and the power stabilization phase of the demand response;

[0105] The power drop phase of the demand response is used to represent the change in the number of air conditioning load start-up state groups and aggregate power over time from time t0 when the control command is issued to time t3 when the first state group in the state queue cools down to the new lower limit of temperature.

[0106] The state group represents the air conditioner within that state unit;

[0107] The power stabilization phase is used to indicate that the air conditioning state group is in a new temperature range [T]. min,2 ,T max,2 During operation, the changes in the number of air conditioning load operating states and aggregated power over time, corresponding to t3 to t4. recv Time; it includes multiple complete cycles;

[0108] Each complete cycle includes: time t3 when the first state group cools down to the new lower limit of temperature, and time t7 when all empty / overlapping state groups have switched to the shutdown state;

[0109] Temperature recovery process after response ends: used to represent the temperature recovery of the air conditioning cluster from a higher temperature range [T] min,2 ,T max,2 Cool down to a lower temperature range [T] min,1 ,T max,1 The process of ].

[0110] like Figure 5 The diagram illustrates the state transition process of the air conditioning load during the load reduction process in the response period. To more intuitively demonstrate the state transition of the air conditioning load over time during the heating process, a continuous curve represents the continuous state queue. The blue curve is formed by connecting the state units of the air conditioning state group that exist at the current moment, and is arranged according to... Figure 2 The numbering order in Figure 5In the continuous state queue, the air conditioning state group numbering from left to right should include (S) off,1 +1) to (S off,1 +S on,1 The temperature drop portion and from 1 to S off,1 The temperature rise portion, that is, while maintaining state continuity, can be considered as the Sth and Sth points of the state queue, respectively. off,1 、(S off,1 +1) Group air conditioners, the state unit numbering method is the same as the state group, and the state unit numbering during operation within the initial temperature range is 1 to (S off,1 +S on,1 ), in the new temperature range numbered 1′~(S off,1 +S on,1 )′ to distinguish them.

[0111] Figure 5 (a) represents the state distribution of the air conditioning load at time t0 when the control command is issued. Figure 5 In (b), t is used. c This indicates the duration for which the air conditioning load participates in temperature control. It can be seen that as the state queue transitions to a higher temperature range, the total number of states remains constant, the number of start-up state groups decreases, and the number of shutdown state groups increases. At this time, the aggregate power is in a decreasing phase, with the following relationship:

[0112]

[0113] t1=t0+min{S on,1 ,ΔF 1,2} (15)

[0114] t2=t0+max{S on,1 ,ΔF 1,2} (16)

[0115] Where t1 represents the polymerization power reaching its minimum amplitude P min At time t2, the polymerization power begins to recover; P agg,1 This represents the aggregate power before the air conditioning load participates in regulation.

[0116] When the Sth off,1 The air conditioning unit lowered the temperature to the new lower limit T. min,2 When the time is t3, then:

[0117] t3=t0+ΔF 1,2 +S on,2 (17)

[0118] The polymerization power will remain relatively stable for a period of time after time t3. Figure 6The load reduction process during the response period, i.e. the number of air conditioner load start-up state groups and the aggregated power change over time during the period t0-t3, is shown.

[0119] Since the temperature control mode changes the on / off time of the air conditioner operation, to analyze the power smooth stage, i.e. the change rule of the aggregated power of the air conditioner load in the new temperature range during the cycle operation after t3, it is necessary to consider that the first and last ends of the state queue reach the minimum indoor temperature T min,2 required time;

[0120] i.e. the S off,1 th group (the first end of the state queue) is warmed up from T max,1 to T max,2 , and then switched to the start-up state to be cooled down to T min,2 , and the required time is (ΔF 1,2 +S on,2 ), and the (S off,1 +1)th group (the tail end of the state queue) is cooled down from T max,1 to T min,1 , and then switched to the off state to be warmed up to T min,2 , and the required time is (S on,1 +S off,1 +ΔF 1,2 -S off,2 .

[0121] As shown in Figure 7 , the size relationship is reflected on the state queue diagram, which includes three cases:

[0122] Case 1: The time when the first state group of the state queue reaches T min,2 is equal to the time when the last state group of the state queue reaches T min,2 ;

[0123] Case 2: The time when the first state group of the state queue reaches T min,2 is greater than the time when the last state group of the state queue reaches T min,2 ;

[0124] Case 3: The time when the first state group of the state queue reaches T min,2 is less than the time when the last state group of the state queue reaches T min,2 ;

[0125] Figure 7 (a) represents the state distribution when the start-stop cycle of the air conditioner after warming up is unchanged, i.e. no state overlap / vacancy phenomenon occurs; Figure 7 (b) represents the state distribution when the start-stop cycle of the air conditioner after warming up is shortened, i.e. the state overlap phenomenon exists, and the air conditioner state has not all entered the new temperature range at t3; Figure 7(c) represents the state distribution in which the on-off cycle becomes longer, i.e. there is a state gap, at (t0+S on,1 + off,1 +ΔF 1,2 - off,2 ) time, all air conditioners have entered the new temperature range.

[0126] Figure 7 (a) the air conditioner load aggregate power in the case shown will remain constant at t3, while the other two cases will cause the aggregate power to periodically fluctuate slightly in the subsequent cycles. In order to facilitate subsequent derivation, let ΔS represent the on-off cycle change before and after the air conditioner load is warmed up, and its expression is:

[0127] ΔS=|S off,1 + on,1 -(S off,2 + on,2 )| (18)

[0128] For Figure 7 (b), from t3, the S off,1 group will switch to the off state, i.e. move to state unit 1', and at the same time, the (S on,1 +ΔS) group will also be warmed up to T min,2 to enter state unit 1', so there will be an overlap of air conditioner state groups (i.e. the same state unit contains two groups of air conditioners); at (t3+ΔS) time, all air conditioners enter the [T min,2 , T max,2 ] temperature range, at this time there are ΔS overlapping state units, and when the overlapping state units pass through the new temperature upper and lower limits, the number of on state groups S on (t) will change in the range [S on,2 , S on,2 +ΔS], causing the aggregate power to periodically fluctuate.

[0129] For Figure 7 (c), at (t3-ΔS) time, all air conditioners enter the [T min,2 , T max,2 ] temperature range, and from t3, the (S off,1 +S on,1 ) state groups of the air conditioner load will circulate in the (S off,2 +S on,2 ) state units with longer periods, and among them there will be ΔS state units without air conditioner groups, which are called gap state units; similar to the overlapping case, when the gap state units pass through the new temperature upper and lower limits, the number of on state groups S on (t) will change in the range [S on,2 -ΔS, S on,2 ], causing the aggregate power to periodically fluctuate.

[0130] Based on the above analysis, we are able to fully model the change in the number of operating states during the load shedding process in the response period:

[0131] First, calculations are performed for some key moments:

[0132]

[0133] Where t4 represents the moment when the first vacant / overlapping state unit reaches the new temperature limit, and before time t4, the vacant / overlapping state units are all distributed between 1′ and S. off,2 In the state unit, starting from time t4+1, the empty / overlapping state units gradually transition to the state unit representing power-on; t5 indicates that the last empty / overlapping state unit arrives at state unit (S). off,2 +1)′ (the moment of switching from power off to power on); t6 and t7 represent the arrival time of the first end of the vacant / overlapping state unit at the state unit (S). off,2 +S on,2 At time t6+1, the empty / overlapping state units gradually transition to the state units indicating shutdown, and by time t7, all empty / overlapping state groups have switched to the shutdown state.

[0134] Before time t3, the number S of air conditioning units that are in operation under the three conditions. on The expressions for (t) are the same, both being...

[0135]

[0136] After time t3, the start-stop cycle of the air conditioning load group changes. The variable J is used to characterize the comparison between the old and new start-stop cycles, and its expression is as follows:

[0137] J = sign(S) off,1 +S on,1 -S off,2 -S on,2 ) (twenty one)

[0138] The Sign() function refers to the sign function in mathematics, where J is -1, and 0 and 1 are respectively associated with the sign function. Figure 7 There are three corresponding scenarios.

[0139] Therefore, the number of air conditioning units that are in operation after time t3 can be represented as follows:

[0140]

[0141] In the above formula, the number of air conditioner operating states changes cyclically starting from time t7+1, with a cycle period of S. off,2+S on,2 It is worth noting that, for cases where the start-stop cycle is longer, the polymerization power begins its first cycle at time t3-ΔS and ends at time t6; for cases where the start-stop cycle is shorter, the polymerization power begins its first cycle at time t3+ΔS and ends at time t7. In this application, t3 is taken as the starting point for the polymerization air conditioner to cycle within the new temperature range, and the period from t3 to the start of the temperature adjustment operation is called the power stabilization stage.

[0142] To maintain the continuity of the state queue and ensure that the aggregated power of the air conditioning load returns to its pre-control level after the temperature setpoint is restored, the callback method needs to be improved similarly to the aforementioned temperature rise control process. This involves selecting an appropriate temperature callback start time and setting different temperature setpoints for the air conditioners in the on and off states, respectively. In the state queue, the first and last ends are respectively the Sth... off,1 Group and (S) off,1 +1) Group, during the callback process, it is necessary to ensure that the Sth group... off,1 The group was the first to cool down to T min,2 Below. When the Sth off,1 The air conditioning unit returns to state unit S. off,1 At this time, S on (t)=S on,1 S off (t)=S off,1 The temperature correction process has ended.

[0143] During the temperature recovery process after the response ends, for air conditioners in the off state, the temperature setpoint is set to T. set,2 This keeps them warm;

[0144] For air conditioners that are turned on, adjustments should be made based on whether there is any overlap in operating status:

[0145] If there is no state overlap, i.e., S off,1 +S on,1 ≤S off,2 +S on,2 When the temperature is adjusted back to the initial temperature setting T, the setpoint will be restored. set,1 This allows each air conditioning status group to orderly transition to its original lower temperature limit;

[0146] If there is a state overlap phenomenon, i.e. S off,1 +S on,1 >S off,2 +S on,2 At that time, directly using temperature control cannot effectively separate already merged overlapping state groups back into their original states. According to... Figure 8(b)It is known that the overlap of state queue refers to the overlap of the head and tail, if the uniform callback temperature setting value is set, the ΔS state groups of the tail will follow the corresponding overlapping state groups of the head to T min,1 transfer, resulting in the temperature diversity of the air conditioner being destroyed after temperature callback, the method adopted by the present application is to separate the overlapping state groups and then perform callback respectively;

[0147] Specifically, at t3, the air conditioners in the (S off,1 +1)~(S off,1 +ΔS) groups are marked as keep groups, and the rest of the air conditioners are classified as normal groups, after the callback operation starts, the air conditioners in the normal groups are allowed to perform the callback operation in the manner of no state overlap, while the keep group air conditioners continue to run in the new temperature range for a full cycle, and then the temperature is adjusted in the manner of the normal group, thereby achieving the effect of separating the overlapping state groups.

[0148] After determining the above scheme, the temperature recovery process after the response ends can be modeled, before modeling, the actual callback start time of the air conditioner load needs to be determined, in order to ensure that the aggregate power of the air conditioner load remains near P agg,1 after temperature callback, each air conditioner needs to complete the same number of complete cycles in the new temperature range before adjusting back, and the earliest callback time t recv of the air conditioner load satisfies:

[0149]

[0150] t recv =t3+l min ×(S on,2 +S off,2 ) (24)

[0151] Where t d is the time when the air conditioner receives the callback signal, and l min represents the minimum cycle number of the air conditioner load running in the new temperature range.

[0152] If t recv is taken as the start time of air conditioner temperature callback, and ΔN 2,1 calculated by formula (12) is combined, the temperature recovery process after the response ends under the condition of no overlapping state is represented as:

[0153]

[0154] For the case of existing overlapping state, the size relationship between ΔN 2,1 and the number of overlapping states ΔS needs to be considered, and the number of air conditioner start-up groups at each key time changes, and the temperature recovery process after the response ends is represented as:

[0155]

[0156] Where t8 represents at t recv After time 1, S remained constant for a period of time. on (t) The moment when the change begins (ΔN) 2,1 (≠ΔS), t9 represents the moment when the power rebound reaches its peak, t 10 This indicates the moment when the aggregate power begins to decline; the head of the state queue (Sth generation) will be... off,1 The air conditioner of the group returned to state unit S. off,1 At that moment, t fin This indicates the completion time of the temperature recovery operation.

[0157] like Figure 8 As shown, in one embodiment of this application, in order to verify the state gap and overlap phenomenon proposed in this application, two groups of aggregation groups with different parameters are selected for simulation analysis. The simulation time interval is 30s. It is assumed that both aggregation groups receive a temperature increase of 1°C at the 15th minute and issue a temperature recovery command 30 minutes later. The control terminal then performs corresponding operations on all air conditioners in each group according to the improved temperature control method.

[0158] Figure 8 (a) and Figure 8 (b) The aggregate power curves of the groups are shown respectively for the state overlap phenomenon caused by the shortening of the start-stop cycle and the state gap phenomenon caused by the lengthening of the start-stop cycle. The specific parameters and calculation results at key moments are shown in Table 1.

[0159] Table 1. Calculation results of key parameters of the state queue model of the two air conditioning aggregation groups.

[0160]

[0161]

[0162] Figure 8 The horizontal axis markings, indicated by black dashed lines, correspond to t0, t3, and t, respectively. recv At that moment. As can be seen, Figure 9 (a) The polymerization power is not from t recv Instead of rebounding from time t8, the temperature remained constant until time t8. This is consistent with the analysis of the temperature recovery process with overlapping states in this application. Previous aggregated power models assumed that the aggregated power of all air conditioners remained constant after entering the new temperature range, without considering the impact of start-stop cycle changes on aggregated power. This resulted in errors in the assessment of the adjustability potential of the air conditioner load group. Obviously, the power curve of the model proposed in this application can better fit the actual situation.

[0163] To meet the power reduction requirement and minimize the power fluctuation during demand response, the application further provides a dispatch optimization method for dispatching and optimizing the dynamic response model of the load group state queue model, which includes:

[0164] The air conditioners are clustered into K aggregated groups, and the user comfort is quantified by predicting the average vote;

[0165] A dispatch optimization model for one-time response of the user is constructed according to the user comfort;

[0166] The dispatch optimization model for one-time response of the user is optimized to obtain a dispatch optimization model for multiple responses of the user.

[0167] Suppose that a load aggregator can control N air conditioners under its jurisdiction on a certain day, which are clustered into K aggregated groups. The load aggregator evaluates the load control potential based on the previous day's outdoor temperature prediction, air conditioner parameters, user temperature setting initial value, and user acceptability information, and reports it to the power company. The power company allocates the control task for the day according to the load prediction result, and requires the load aggregator to provide a certain control capacity P need according to demand response. The load aggregator formulates a dispatch plan for each aggregated group accordingly.

[0168] When formulating the dispatch plan, the load aggregator needs to try to ensure the user comfort experience. Thermal comfort is the subjective evaluation and feeling of the user on the indoor thermal environment. To quantify the influence of temperature on human thermal comfort, the application uses the predicted mean vote (PMV) index to measure the user's thermal comfort;

[0169] ISO7730 recommends that the PMV value be between ±0.5, so when all user parameters except indoor temperature are known, the temperature range that each user can accept can be deduced according to the calculation formula of the PMV index For most residential air conditioners, the temperature setting value is adjusted in units of 0.5℃, so the application sets 0.5℃ as the minimum unit of temperature adjustment.

[0170] The temperature adjustment set of the i-th air conditioner and the maximum temperature adjustment of the aggregated group I (I∈1, 2, …, K) can be expressed as follows:

[0171]

[0172] where Z represents the set of integers, represents the maximum temperature adjustment that the user of the i-th air conditioner can accept, represents the expected temperature set value of the user of the i-th air conditioner, in this application, the temperature set initial value of the user i is considered is equal to .

[0173] Based on the classification modeling of the aggregated power change process of the air conditioner load after the response control, the heat capacity C I of the indoor gas of the aggregated group, the thermal resistance R I of the indoor gas, the air conditioner refrigeration efficiency η I and the rated power key parameters are obtained through the load aggregator, and then the number of on-off machine state groups before and after the temperature increase is calculated according to the temperature increase amount allocated by the group and the corresponding and and the number of air conditioner groups participating in the regulation and control regulation start and end time and are taken as the decision variables of the scheduling optimization, so as to construct the scheduling optimization model of the user one-time response.

[0174] The scheduling optimization model of the user one-time response includes a first stage and a second stage;

[0175] The first stage schedules and optimizes the load reduction process during the response period, and the optimization goal is to minimize the deviation between the actual load reduction amount and the demand amount;

[0176] Specifically, the load aggregator finds the optimal combination of the starting time temperature increase amount and the number of participants variables of each aggregated group participating in the regulation and control, so that the deviation between the actual load reduction amount ΔP agg (t) and the demand amount P need is minimized, and the objective function is:

[0177] minF1=Var(ΔP agg (t)-P need ),t start ≤t≤t end (29)

[0178]

[0179] Where t start and t end represent the start and end times of the demand response period, and in the optimization process of the first stage, each aggregated group is defaulted to be at t endFollowing the improved method, a temperature callback operation was implemented. Based on the above response process, the number of startup states for each aggregation group was deduced. The relationship between it and its variables is simplified by functions f1 and f2. In equation (31), Son and Soff can be calculated according to equations (9) and (10). The results are calculated according to equations (11) and (12), where ts is a variable and tend is a given value. The variable relationship in (9) to (12) corresponds exactly to f2 in equation (32).

[0180] The constraints for the first stage are:

[0181]

[0182] The second stage involves scheduling and optimizing the temperature recovery process after the response ends. The optimization objective is to keep the polymerization power during the temperature recovery process stable within the rebound load limit.

[0183] Since users' disorderly reduction of temperature setpoints after demand response will inevitably lead to a significant load rebound, endangering power grid safety, this application adds a delay unit to each group during the temperature recovery phase: allowing each aggregation group to maintain its current temperature setpoint for several cycles before adjusting the temperature back.

[0184] During the temperature recovery phase, the polymerization power remains stable within the rebound load limit. The objective function for the second phase is:

[0185]

[0186] The second-stage constraints are:

[0187]

[0188] in, This indicates that the I-th aggregation group optimized the temperature callback timing. Number of power-on states at time t after the current state. The moment when the last group of air conditioners returns to its original temperature state can be calculated according to equations (25) and (26). At this time, the aggregate power will return to the initial power P. agg,1 β represents the percentage of rebound load fluctuation.

[0189] After completing the scheduling optimization model for a one-time user response, based on previous temperature control studies, each air conditioner can only participate in at most one response process, and temperature reversal is not allowed during the time from entering the control to the end of the response. This mode can easily lead to significant differences in the comfort experience among users due to the different control durations of the air conditioners selected at different times.

[0190] To solve the above problems, a scheduling optimization model is proposed to allow users to participate in demand response multiple times, and the following improvements are made based on the scheduling optimization model for one-time response of the user proposed in the application:

[0191] Allowing the air conditioner to end the response in advance after temperature control;

[0192] Limiting the number of adjustments of each aggregation group;

[0193] Limiting the number of air conditioners in a controlled state at any time to be no more than the total number of air conditioners in the aggregation group;

[0194] Prioritizing the priority of air conditioner users participating in the response.

[0195] Since the air conditioner is allowed to end the response in advance after temperature control in this control mode, the control variable of the first stage should include the temperature adjustment start time Second, considering user comfort experience, while reducing the calculation pressure of the load aggregator when formulating the control strategy and the communication pressure when implementing the control, the number of adjustments of each aggregation group is limited to n ctrl times.

[0196] Correspondingly, the decision variables such as control start and end time, temperature adjustment amount, and the number of participating air conditioners are converted into the following n ctrl dimensional vectors, and the objective function F1 still satisfies equation (29);

[0197]

[0198] To ensure that the selected air conditioner can only participate in one response scheme at any time, i.e., the number of air conditioners in a controlled state at time t is no more than the total number of air conditioners in the aggregation group, the constraint condition (33) should be replaced by:

[0199]

[0200] Where, represents the number of air conditioners extracted by the aggregation group in the jth control, and then the start and end times of the jth control correspond respectively.

[0201] Based on the optimization results of the demand response stage, the second stage only needs to optimize the temperature adjustment time of the last control of each aggregation group, and the objective function and constraint conditions remain the same as the basic scheduling model, which will not be repeated here.

[0202] Since the load aggregator issues the control instruction to each aggregation group once according to the day-ahead dispatching plan (assuming that the demand capacity is not changed once it is determined, and the power shortage caused by unexpected events will be responded by the power company through other measures), but before the control strategy is implemented during the day, the load aggregator needs to select the air conditioner users of each aggregation group participating in the regulation:

[0203] First, the distribution of the air conditioner temperature state under the parameter condition of each aggregation group is calculated through formula (9) (10), and then the air conditioners with similar indoor temperatures are divided into the same state group, and these state groups are one-to-one corresponding to state units.

[0204] In a regulation process, the load aggregator uniformly selects air conditioners from each state group to participate in the response. Under normal circumstances, the incentive cost of users participating in demand response in the same aggregation group is similar, and it can be considered that the probability of selecting the air conditioner users in each state group is the same.

[0205] In order to ensure that the users in the same aggregation group are affected to a similar degree, the comfort factor is used as an index to guide the load aggregator to sort the priority of the air conditioner users participating in the response. The comfort factor uses the product of the temperature adjustment change of the air conditioner and the controlled time to represent the degree of influence of the user comfort, and the expression of the comfort factor C i (t) of each user in the aggregation group I at the current time is:

[0206]

[0207] The smaller the result is, the smaller the degree of influence of the comfort of the user i before the jth regulation is, and the higher the priority of the air conditioner selected is.

[0208] In an embodiment of the present application, in order to verify the effectiveness of the aggregation power dynamic response model, the k-means clustering method is first used to preliminarily group the parameter heterogeneous air conditioner cluster, and the clustering features are usually selected as the temperature change parameter RC and the characteristic temperature difference QR of the controlled air conditioner. However, in fact, the aggregation power curves of the same homogeneous air conditioner with similar parameters are quite different even if the same temperature is increased when the initial temperature setting value is different, as shown in Figure 9 .

[0209] Figure 10By using the improved temperature control method to simulate and control 500 homogeneous air conditioners, it is obtained that different curves respectively represent the aggregated power when the user expected temperature (temperature setting initial value) is 25.5℃, 26℃ and 26.5℃, the demand response period is from 30 min to 110 min, and the temperature up-regulation is 1℃. Combined with formulas (8)-(10), the state queue distribution (on-off state duty cycle) of the homogeneous air conditioner under different temperature setting values is different, and the higher the temperature setting initial value of the air conditioner, the lower the initial level of the aggregated power. In addition, even if the temperature up-regulation is 1℃, but the change amount of the start-stop cycle of the air conditioner increases with the increase of the temperature setting initial value, and the more obvious "state loss" phenomenon occurs, resulting in more intense power fluctuation in the power stable stage.

[0210] In view of the above, 1500 homogeneous air conditioners are controlled in two ways: a. without further grouping, and controlled according to the temperature setting initial value of 26℃; b. divided into 3 groups according to the temperature setting initial value, and controlled according to the 3 temperature setting initial values. The response start and end time of each group is still set to 30 min and 110 min, and the aggregated power curve is as shown in Figure 11 It can be seen that directly controlling the air conditioner clusters with different temperature setting initial values uniformly will destroy the continuity of the state queue and the temperature diversity, resulting in the red curve keeping cyclic fluctuation of the aggregated power after the temperature is adjusted back. If the air conditioners are further grouped according to the temperature setting initial value, on the one hand, the aggregated power can be kept stable after the temperature is adjusted back, and on the other hand, due to the difference in start-stop cycle, there is a time difference between the actual temperature adjustment time of each group after the response end instruction is issued, which to some extent relieves the load rebound caused by the simultaneous temperature down-regulation of the air conditioners.

[0211] In addition, the adjustable range of the air conditioner with different temperature setting initial values is different when receiving temperature control, for example, the higher the temperature setting initial value, the smaller the temperature up-regulation that the user can accept, so the load reduction amount provided by each air conditioner in the same aggregated group is uneven. Combined with formula (28), if the users are not distinguished according to the temperature setting initial value, in the mode of unified participation of the aggregated group, it is not conducive to exert the control potential of the air conditioner load. In summary, the temperature setting initial value should be an important reference factor for clustering and grouping of the air conditioner load.

[0212] In an embodiment of the present application, it is assumed that more than 12000 air conditioners in a certain area (residential area or office building) can participate in demand response in the control period, and according to the user data (including expected temperature setting value, temperature adjustable range, air conditioner parameters, etc.) collected in advance, 20 aggregated groups are divided to accept load aggregation dispatching, and the air conditioner and simulation parameters are as shown in Table 2.

[0213] Table 2 Air conditioner indoor parameters and simulation parameters

[0214]

[0215] On a certain day, the power company needs the region to provide a load shedding amount of proportion a during the period of 13:00-15:00 to alleviate the power supply pressure, and requires that the rebound load percentage β does not exceed 2a, and the load recovery time does not exceed 16:00. In order to verify the superiority and effectiveness of the demand response scheduling strategy proposed in this paper, the following several scheduling strategies that the load aggregator may adopt are compared:

[0216] Scheme 1: Adopt the traditional temperature control method, and simultaneously increase the temperature set value at 13:00 and restore the temperature set value at 15:00 for each aggregation group.

[0217] Scheme 2: Adopt the improved temperature control method to realize the separation control of the on-air conditioner and the off-air conditioner. Under this scheme, the air conditioners of each group still simultaneously increase and restore the temperature at the start and end time of the control task.

[0218] Scheme 3: Based on scheme 2, a control preparation time is introduced to guide each group to enter the temperature control process at different times before the response period to reduce the impact of the load drop process on the power grid. This scheme requires that all air conditioners participating in the control have entered the new temperature range by 13:00. In this paper, the preparation time is set to 1h.

[0219] Scheme 4: A one-time response scheduling optimization model of users, that is, from 13:00, 15min is taken as an optimization time window, and part of the air conditioners of different groups are guided to enter the temperature control process at different times in each time window, but the temperature is not allowed to be restored before 15:00. The aggregated power reduction of each time window is provided by two parts:

[0220] (1) The rapid power drop caused by the start of the response process of part of the air conditioners in the current time window;

[0221] (2) The air conditioners that have completed the temperature rising process are in the power stable stage in the current time window, but can still provide certain power reduction due to the decrease of the on-off duty ratio, and the power complementary characteristics of the controlled air conditioners among the aggregation groups and in each time window are comprehensively utilized through appropriate optimization combination to meet the control demand.

[0222] Scheme 5: A multi-time response scheduling optimization model that allows users to participate in control or restore the temperature according to the improved temperature control method at any time from 13:00 to 15:00, and can participate in response again after the temperature is restored. The temperature increase amount and the number of air conditioners of each aggregation group are independent of each other.

[0223] Figure 12(a) and (b) can intuitively show the change of the group aggregation power of a certain aggregation group in the embodiments 4 and 5, and the demand response stage is from the 60th minute to the 180th minute. In the scheme 4, two batches of air conditioners are selected to participate in the response in the first and third time windows, the load reduction amount of each control is superimposed on each other, and the air conditioner load is adjusted back in time after the end of the response period; in the scheme 5, the air conditioner participating in the control is allowed to enter the secondary control after the temperature is recovered, and the temperature curve of the air conditioner continuously participating in the control twice is shown by the red dotted line, wherein the temperature up-regulation amount of the two controls is 0.5°C and 1°C respectively, and the group aggregation power appears twice drop and twice rebound.

[0224] In the simulation experiment of the schemes 4 and 5, considering that the temperature control mode is adopted in the present application to ensure temperature diversity, a large number of air conditioners cannot complete the on-off state switching in a short time to achieve the effect of a large instantaneous drop of the aggregation power, so the aggregation group is allowed to enter the temperature adjustment process 10 minutes in advance as a preparation stage of the demand response.

[0225] Table 3 shows the load reduction deviation and rebound fluctuation percentage β of the above-mentioned five scheduling schemes when the power down-regulation target proportion α is 5%, 10%, and 15% respectively:

[0226] Table 3 shows the load reduction deviation and rebound load percentage of each scheduling scheme

[0227]

[0228] Since the aggregation power fluctuation of the schemes 1 and 2 is large, the present embodiment will mainly further analyze the scheduling optimization results of the schemes 3, 4, and 5.

[0229] In combination with Table 1 and Figure 13 The schemes 4 and 5 have more advantages in the load reduction capacity and rebound suppression effect, and the two schemes make the air conditioners participate in the temperature control process in batches at different time points in the demand response period, thereby avoiding the power valley phenomenon in the schemes 1 to 3.

[0230] The scheme 3 is improved based on the scheme 2, and the essence is to concentrate the power drop stage of each aggregation group participating in the temperature control to before the control task starts, so as to ensure that each group is in the power stable stage after the temperature control at 13:00, and the aggregation total power is maintained near the control target. Since the preparation time of 1h is introduced, the scheme to a certain extent alleviates the situation that all air conditioners up-regulate the temperature at the same time, but obviously the aggregation power will still drop by a large amplitude before 13:00, which causes impact on the power grid operation. In addition, the scheme depends on the load reduction amount caused by the temperature setting value rising and the on-off ratio of the air conditioner decreasing, and the load reduction amount is limited, and with the down-regulation proportion increasing from 5% to 15%, the aggregation power in the response period will be difficult to maintain near the target power, and the load fluctuation will be more intense.

[0231] Compared with scheme 3, scheme 4 is based on the dynamic response model of the air conditioner load group state queue model of the present application, and comprehensively utilizes the power reduction generated in the two stages of power drop and power stabilization of each aggregation group.

[0232] Figure 5 The group aggregation power comparison of the 5th aggregation group under the 10% reduction ratio when implementing schemes 3 and 4 is shown, the response start and end time, temperature up-regulation amount and the number of participating air conditioners of the two schemes are shown in Table 4, the user average load reduction amount index ΔP is defined to measure the load reduction capacity of the air conditioners in the group when implementing different schemes, and the calculation formula is:

[0233]

[0234] Among them, P5 represents the aggregation power of the 5th group air conditioner without implementing any scheme.

[0235] Table 4 Scheduling results of the 5th group air conditioner when implementing schemes 3 and 4

[0236]

[0237] From Table 4, it can be seen that the power provided by each air conditioner in the aggregation group when implementing scheme 4 is 16W higher than that of scheme 3 when the temperature is up-regulated by 1℃, which shows that the control mode of allowing air conditioners to enter the response multiple times and in batches can better exert the regulation potential of air conditioner load.

[0238] Further considering the difference in thermal comfort of air conditioner users, this paper proposes an improved scheme 5, from Table 3, it can be seen that the ability of schemes 4 and 5 to control load reduction deviation and suppress rebound load is generally similar, the difference in the effects of the two schemes mainly reflects on the comfort factor of each user during the demand response period, and Table 5 shows the user comfort factor calculation results of implementing schemes 4 and 5.

[0239] Table 5 Comfort factor index of air conditioner users under different reduction ratios when implementing schemes 4 and 5 respectively

[0240]

[0241] From ​It can be seen that, as the target down-regulation ratio increases, the comfort level of the controlled air conditioner users is gradually affected, and the load aggregator needs to select more air conditioners to participate in demand response in order to provide sufficient power reduction. According to the average value and variance of the comfort factor, it can be seen that scheme 5 reduces the total comfort factor of the air conditioner cluster to a certain extent, and also reduces the comfort difference between users. The reason is that scheme 5 preferentially selects the air conditioner with the smallest current user comfort level to participate in temperature control, and the power reduction pressure is evenly distributed to the air conditioner load group, avoiding the extreme situation that only part of the air conditioners are continuously controlled at high temperature for a long time to provide load reduction, and the remaining air conditioners do not participate in the demand response process at all.

[0242] The above examples are only used to illustrate the technical solutions of the present application, but not to limit them; although the present application has been described in detail with reference to the foregoing examples, those skilled in the art should understand that they can still modify the technical solutions recorded in the foregoing examples, or make equivalent replacement for part of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. An air conditioning load cluster control and scheduling optimization method, characterized in that, The application relates to a method for scheduling optimization of an air conditioner load group state queue model. According to the equivalent thermal parameter model of a single air conditioner, an air conditioner load state queue model is constructed. The air conditioner load state queue model divides the start-stop cycle of the air conditioner into , state units according to time intervals t off + t on . Wherein, the first state unit indicates that the air conditioner compressor is in a shutdown state, and the indoor temperature gradually increases. t off The second state unit indicates that the air conditioner is in a startup cooling state. t on The second state unit indicates that the air conditioner is in a startup cooling state. every time interval the air conditioning system in each state unit is transferred to the next state; According to the influence of temperature control on the aggregated power of air conditioners, a dynamic response model of the air conditioner load group state queue model is constructed. The dynamic response model of the air conditioner load group state queue model divides the change process of the aggregated power of air conditioners after response control into three stages, i.e. a load reduction process during response, a temperature recovery process after response and a power recovery process after the temperature recovery process. The state group represents air conditioners in each state unit. The power drop phase of the demand response is used to represent the time from the dispatch of the regulation instruction to the first state group of the state queue cooling down to the lower limit of the new temperature t 0 moment to the first state group of the state queue cooling down to the lower limit of the new temperature t 3 moment, the number of air conditioning load start state groups and the change of the aggregated power with time During the temperature recovery process after response, the temperature set value of an air conditioner in the shutdown state is set as a temperature set value after temperature rise. The power smoothing stage is used to represent the case that the number of air conditioning load start state groups and the aggregated power change over time when the air conditioning state group operates in the new temperature range T min,2 , T max,2 ] corresponds to t 3 to the earliest temperature rollback time allowed for the air conditioning load t recv ; it includes multiple complete cycles; The power smoothing stage reaches according to the head and tail of the state queue respectively The correlation of the time required is modeled, which includes three cases: Case 1: No state overlap or gap occurs, the first state group of the state queue arrives at the time equal to the time the last state group of the state queue arrives . Case 2: When there is state overlap, the first state group of the state queue arrives at a time greater than the time at which the last state group of the state queue arrives . Case 3: In case of state vacancy, the first state group of the state queue arrives at a time less than the time at which the last state group of the state queue arrives at a time less than the time at which the last state group of the state queue arrives wherein represents the minimum value of the indoor temperature after the temperature is raised; Each complete cycle includes: the first group of states is cooled to the lower limit of the new temperature at time t 3 when all vacant / overlapping groups of states have been switched to the off state t 7 time; Temperature recovery process after response end: a process for indicating that the temperature of the air conditioning cluster is lowered from a higher temperature range[ T min,2 , T max,2 ] to a lower temperature range[ T min,1 , T max,1 ] For an air conditioner in the startup state, the temperature set value is adjusted according to whether there is state overlap. If there is no state overlap, the temperature set value is set as a temperature set value before temperature rise. If there is state overlap, the overlapped state groups are separated and then adjusted. According to the power characteristics of the dynamic response model and the comfort requirements of users, scheduling optimization of the aggregated air conditioner load participating in the demand response process is completed. The method for scheduling optimization of the dynamic response model of the air conditioner load group state queue model comprises the following steps.

2. The air conditioning load clustering control and dispatch optimization method of claim 1, wherein, The number of air conditioners contained in each state group in the air conditioner load state queue model is ( t off + t on) ; wherein, N is the total number of air conditioners.

3. The air conditioning load clustering control and dispatch optimization method of claim 1, wherein, A scheduling optimization model of user one-time response is constructed according to user comfort. The air conditioners are clustered into K aggregated groups, and user comfort is quantified by predicting average votes. The scheduling optimization model of user one-time response is optimized to obtain a scheduling optimization model of user multiple-time response. The scheduling optimization model of user one-time response comprises a first stage and a second stage.

4. The air conditioning load clustering control and dispatch optimization method of claim 3, wherein, The first stage schedules and optimizes the load reduction process during response, and the optimization target is to minimize the deviation between the actual load reduction amount and the demand amount. The second stage schedules and optimizes the temperature recovery process after response, and the optimization target is to keep the aggregated power of the temperature recovery process within the rebound load limit and keep the power stable. The optimization of the scheduling optimization model of user one-time response comprises the following steps.

5. The air conditioning load clustering control and dispatch optimization method of claim 3, wherein, The temperature-controlled air conditioner is allowed to end the response in advance. The number of times of adjustment of each aggregated group is limited. The number of air conditioners in the controlled state at any time is limited to be not more than the total number of air conditioners in the aggregated group. The priority of air conditioner users participating in response is sorted. ​

Images