Cooperative regulation and control method and system for air conditioner load virtual synchronous machine cluster

Abstract

The invention relates to the technical field of air conditioner load regulation and control, in particular to an air conditioner load virtual synchronous machine cluster cooperative regulation and control method and system. The system comprises a data acquisition module, a candidate terminal screening module, a dynamic grouping module, a virtual synchronous machine modeling module, an adjusting power calculation module, an instruction decomposition and issuing module and a data storage and monitoring module. The method comprises the following steps: firstly, introducing a dynamic clustering algorithm based on real-time running state characteristics of air conditioner loads, identifying the running state similarity of the loads in real time, and dividing time-varying clusters; then, a personalized virtual synchronous machine model is constructed for each air conditioner cluster, inertia and damping characteristics of a simulated generator are endowed to the air conditioner cluster, and the air conditioner cluster can effectively participate in inertia support and frequency modulation response of a power grid; and finally, a cooperative regulation and control system is constructed, and autonomous, rapid and orderly response of the air conditioner cluster to the power grid frequency change is achieved.

Description

Technical Field

[0001] This invention relates to the field of air conditioning load control technology, specifically to a method and system for coordinated control of air conditioning load virtual synchronous machine clusters. Background Technology

[0002] The increasing penetration of intermittent power sources such as wind and solar power in the power system has led to more severe frequency fluctuations and a decrease in inertia. Traditional methods of providing frequency regulation services relying on thermal power and energy storage face problems such as high costs and limited capacity. Meanwhile, air conditioning accounts for 30%-50% of the building load, possessing enormous adjustability potential and considered a high-quality flexible load resource.

[0003] Existing air conditioning load control technologies mainly suffer from the following problems:

[0004] First, most control schemes adopt static grouping strategies, failing to consider the dynamic changes in terminal operating status, making it difficult to form time-varying clusters to adapt to real-time fluctuations in grid frequency. Second, there is a lack of virtual synchronous machine modeling methods for air conditioning load clusters, making it impossible to effectively simulate system inertia and damping characteristics, and difficult to participate in grid inertia support and frequency regulation response. Third, traditional clustering methods for dividing air conditioning groups cannot adapt to the time-varying nature of ambient temperature and user behavior patterns. Fourth, the data storage and monitoring mechanisms during the control process are imperfect, lacking real-time tracking of cluster operating status and control effects, making it difficult to support the optimization and iteration of subsequent control strategies.

[0005] To address the aforementioned issues, it is necessary to propose a collaborative control method and system for air conditioning load virtual synchronous machine clusters. By introducing a dynamic clustering algorithm to identify the similarity of the operating states of air conditioning loads, and constructing a virtual synchronous machine model for each cluster, it is endowed with the inertia and damping characteristics of a generator. This enables the air conditioning clusters to respond autonomously, rapidly, and orderly to changes in grid frequency, thereby improving the stability and resilience of the power distribution system while ensuring user comfort. Summary of the Invention

[0006] To address the problems in related technologies, this invention proposes an integrated energy system for zero-carbon industrial parks to overcome the aforementioned technical issues in existing technologies. The purpose of this invention is to first introduce a dynamic clustering algorithm based on the real-time operating status characteristics of air conditioning loads to identify the similarity of load operating states and classify them into time-varying clusters; then, to construct a personalized virtual synchronous machine model for each air conditioning cluster, endowing it with the inertia and damping characteristics of a generator, enabling it to effectively participate in the inertia support and frequency regulation response of the power grid; finally, to construct a collaborative control system to achieve autonomous, rapid, and orderly response of the air conditioning clusters to changes in power grid frequency.

[0007] To achieve the above objectives, the present invention provides the following technical solution: a method for coordinated control of a virtual synchronous machine cluster for air conditioning load, comprising the following steps:

[0008] Step 1: Collect operating parameters from multiple air conditioning terminals every 15 minutes, filter out adjustable candidate air conditioning terminals, and construct a multi-dimensional data vector for each candidate air conditioning terminal;

[0009] Step 2: Use the multidimensional data vector to dynamically group the air conditioning terminals to form several time-varying clusters;

[0010] Step 3: Construct a virtual synchronization machine model for each cluster;

[0011] Step 4: Based on the power grid frequency deviation signal, calculate the regulation power that each cluster should provide using the virtual synchronous machine model;

[0012] Step 5: Decompose the adjustment instructions to the individuals within the cluster and issue them for execution.

[0013] Preferably, in step one, the operating parameters of the air conditioning terminal include the indoor temperature T. in Set temperature T set Power consumption P, building thermal time constant τ, ambient temperature T out ;

[0014] The multidimensional data vector is:

[0015] x i =[T in ,T set ,|T in -T set |,P,T out ,τ]

[0016] Where, x i Let i represent the multidimensional data vector of the i-th air conditioner.

[0017] Preferably, the specific steps of step two are as follows:

[0018] Step 2.1: Normalize the multidimensional data vector;

[0019] Step 2.2: Calculate the weighted normalized Euclidean distance between the air conditioning terminals sequentially. The calculation formula is as follows:

[0020]

[0021] Where, d ij ω represents the weighted normalized Euclidean distance between the i-th air conditioning terminal and the j-th air conditioning terminal. k σ represents the weight coefficient of the k-th dimension feature.k The standard deviation represents the k-th dimension feature;

[0022] Step 2.3, set the distance threshold to θ, if d ij If ≤θ, it means that the i-th air conditioning terminal and the j-th air conditioning terminal have similar operating characteristics; otherwise, it means that they are not similar.

[0023] Step 2.4: Count the number of air conditioning terminals with similar operating characteristics to each of the aforementioned air conditioning terminals, and sort them in descending order of quantity;

[0024] Step 2.5: Group the top-ranked air conditioning terminal and all air conditioning terminals with similar operating characteristics into cluster G. m ;

[0025] Step 2.6: Repeat steps 2.4 to 2.5 for the air conditioning terminals that are not assigned to a cluster, until all the air conditioning terminals are assigned to a cluster.

[0026] Preferably, in step three, the calculation formula for the virtual synchronizer model parameters is as follows:

[0027]

[0028] K m =γ·η m

[0029]

[0030] Among them, J m D represents the equivalent moment of inertia. m K represents the equivalent damping coefficient. m C represents droop gain or response sensitivity. i G represents the heat capacity of the i-th air conditioner. m Let |G| represent the m-th air conditioning cluster. m | represents the number of air conditioning terminals in the m-th air conditioning cluster, η m T represents the average regulation margin of the m-th air conditioning cluster, α, β, and γ represent coefficients. in,i Indicates the current indoor temperature, T low Indicates the minimum comfortable temperature, T high P represents the highest comfortable temperature. i P represents the current operating power. rated This indicates the rated maximum power.

[0031] Preferably, in step four, each cluster should provide the following formula for calculating the adjustment power:

[0032]

[0033] Where, ΔP m Δf represents the regulation power required for the m-th air conditioning cluster, and Δf represents the power grid frequency deviation signal.

[0034] Preferably, the specific steps of step five are as follows:

[0035] Step 5.1: Calculate the comprehensive score for each air conditioning terminal in the cluster sequentially. The calculation formula is as follows:

[0036] S i =ω1F1+ω2F2+ω3F3

[0037]

[0038] Among them, S i F1 represents the overall score of the i-th air conditioning terminal, F2 represents the temperature margin score, F3 represents the regulation capability score, and F4 represents the response speed score. ω1, ω2, and ω3 represent the average response time, and ω1, ω2, and ω3 represent the score weighting coefficients.

[0039] Step 5.2: Sort the air conditioning terminals by their comprehensive scores from highest to lowest, select them sequentially, and accumulate their adjustable power until the cumulative adjustment capacity is ≥ ΔP. m ;

[0040] Step 5.3: Distribute the adjustment amount obtained by each air conditioning terminal and execute it.

[0041] To achieve the above objectives, the present invention also provides the following technical solution: an air conditioning load virtual synchronizing machine cluster collaborative control system, which applies the above-mentioned air conditioning load virtual synchronizing machine cluster collaborative control method, comprising:

[0042] The data acquisition module is used to acquire the operating parameters of multiple air conditioning terminals and the power grid frequency deviation signal, and to preprocess the acquired data.

[0043] The candidate terminal screening module is used to screen adjustable candidate air conditioning terminals in the cluster based on the preprocessed operating parameters, exclude terminals that have just completed a start-up and shutdown, whose room temperature exceeds the comfortable temperature range, whose equipment is faulty or offline, and construct a multi-dimensional data vector for each candidate terminal.

[0044] The dynamic grouping module is used to evaluate the similarity between candidates based on their multidimensional data vectors by calculating the weighted normalized Euclidean distance between them, and dynamically divide the highly similar terminals into several time-varying clusters.

[0045] The virtual synchro modeling module is used to build a virtual synchro model for each time-varying cluster. It calculates the equivalent moment of inertia, equivalent damping coefficient, and droop gain of the model based on the heat capacity, number, and average regulation margin of the terminals in the cluster.

[0046] The regulation power calculation module is used to receive the grid frequency deviation signal and calculate the regulation power that each cluster should provide through a virtual synchronous machine model.

[0047] The instruction decomposition and distribution module is used to calculate the comprehensive score of each terminal in the cluster, select terminals according to the score and accumulate the adjustable power, decompose the adjustment instructions to the selected individual terminals and distribute them for execution;

[0048] The data storage and monitoring module is used to store the operating data, terminal status, and control results of each module, monitor the cluster's operating status and control effects in real time, and provide data support for subsequent optimization.

[0049] Compared with the prior art, the beneficial effects of the present invention are:

[0050] (1) This invention is a method and system for coordinated control of virtual synchronous machine clusters of air conditioning loads. By introducing a dynamic clustering algorithm to identify the similarity of the operating status of air conditioning loads, and combining real-time updated multidimensional data vectors for clustering, the cluster structure can be adjusted in real time according to changes in ambient temperature, user-set temperature and equipment status, which significantly improves the accuracy and timeliness of cluster division.

[0051] (2) The present invention is a method and system for coordinated control of virtual synchronous machine clusters of air conditioning load. It constructs a virtual synchronous machine model for each cluster, which effectively makes up for the lack of system inertia caused by the grid connection of intermittent power sources such as wind power and photovoltaic, enhances the frequency stability and anti-interference capability of the power grid, and provides reliable inertia support and frequency regulation services for the power grid.

[0052] (3) This invention is a method and system for coordinated control of virtual synchronous machine clusters of air conditioning load. The precise decomposition strategy of the control command takes into account both the power grid demand and user comfort. While meeting the power demand of cluster regulation, it maximizes the comfort of the user's indoor environment and avoids the problem of decreased user experience. This invention fully explores the flexible adjustment potential of building air conditioning load, reduces the economic cost and carbon emissions of power grid frequency regulation, helps to achieve dual carbon goals, and promotes the transformation of the power system towards a clean, low-carbon, safe and efficient direction. Attached Figure Description

[0053] Figure 1 This is a flowchart illustrating the method of the present invention. Detailed Implementation

[0054] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention.

[0055] Example 1

[0056] Please see Figure 1This invention proposes a method for coordinated control of air conditioning load virtual synchronous machine clusters, comprising the following steps:

[0057] Step 1: Collect operating parameters from multiple air conditioning terminals every 15 minutes, filter out adjustable candidate air conditioning terminals, and construct a multi-dimensional data vector for each candidate air conditioning terminal; specifically, for example, air conditioning terminals that have just completed a start-stop cycle can be excluded to prevent damage from frequent compressor start-stop cycles; air conditioning terminals whose room temperature has exceeded the comfortable temperature range, equipment malfunctions, or are offline can be excluded.

[0058] Step 2: Use the multidimensional data vector to dynamically group the air conditioning terminals to form several time-varying clusters;

[0059] Step 3: Construct a virtual synchronization machine model for each cluster;

[0060] Step 4: Based on the power grid frequency deviation signal, calculate the regulation power that each cluster should provide using the virtual synchronous machine model;

[0061] Step 5: Decompose the adjustment instructions to the individuals within the cluster and issue them for execution.

[0062] In this embodiment, the first step of the screening process includes the following criteria: ① Whether it is running; those that are not running are excluded; ② Check the running time; those that have just been turned on or have a running time of less than 10 minutes are excluded; ③ Those whose room temperature is not within a comfortable range are also excluded; ④ Those that are offline, have communication failures, or cannot collect data are excluded because they cannot be controlled.

[0063] Furthermore, in step one, the operating parameters of the air conditioning terminal include the indoor temperature T. in Set temperature T set Power consumption P, building thermal time constant τ, ambient temperature T out ;

[0064] The multidimensional data vector is:

[0065] x i =[T in ,T set ,|T in -T set |,P,T out ,τ]

[0066] Where, x i Let i represent the multidimensional data vector of the i-th air conditioner.

[0067] Furthermore, the specific steps of step two are as follows:

[0068] Step 2.1: Normalize the multidimensional data vector;

[0069] Step 2.2: Calculate the weighted normalized Euclidean distance between the air conditioning terminals sequentially. The calculation formula is as follows:

[0070]

[0071] Where, d ij ω represents the weighted normalized Euclidean distance between the i-th air conditioning terminal and the j-th air conditioning terminal. k σ represents the weight coefficient of the k-th dimension feature. k The standard deviation represents the k-th dimension feature;

[0072] Step 2.3, set the distance threshold to θ, if d ij If ≤θ, it means that the i-th air conditioning terminal and the j-th air conditioning terminal have similar operating characteristics; otherwise, it means that they are not similar.

[0073] Step 2.4: Count the number of air conditioning terminals with similar operating characteristics to each of the aforementioned air conditioning terminals, and sort them in descending order of quantity;

[0074] Step 2.5: Group the top-ranked air conditioning terminal and all air conditioning terminals with similar operating characteristics into cluster G. m ;

[0075] Step 2.6: Repeat steps 2.4 to 2.5 for the air conditioning terminals that are not assigned to a cluster, until all the air conditioning terminals are assigned to a cluster.

[0076] In this embodiment, in step 2.2, the initial weights can be assigned based on expert experience to reflect the importance of different features in the clustering task, such as the indoor temperature T. in The value is 0.8, and the set temperature T is... set =0.7, |T in -T set | is 1.0, power consumption P is 0.6, building thermal time constant τ is 0.5, and external ambient temperature T out It is 0.9.

[0077] In step 2.3, θ is initially set based on experience and then dynamically adjusted. The adjustment depends on the ratio of the air conditioners in the top-ranked air conditioner cluster in the previous round to the set ratio. For example, if the set ratio is 40%, if it is higher than 40%, the value of θ will be reduced in the next round; if it is lower than 40%, the value of θ will be increased in the next round.

[0078] Furthermore, in step three, the calculation formula for the virtual synchronizer model parameters is as follows:

[0079]

[0080] K m =γ·η m

[0081]

[0082] Among them, J m D represents the equivalent moment of inertia. m K represents the equivalent damping coefficient. m C represents droop gain or response sensitivity. i G represents the heat capacity of the i-th air conditioner. m Let |G| represent the m-th air conditioning cluster. m | represents the number of air conditioning terminals in the m-th air conditioning cluster, η m T represents the average regulation margin of the m-th air conditioning cluster, α, β, and γ represent coefficients. in,i Indicates the current indoor temperature, T low Indicates the minimum comfortable temperature, T high P represents the highest comfortable temperature. i P represents the current operating power. rated This indicates the rated maximum power.

[0083] Furthermore, in step four, each cluster should provide the following formula for calculating the adjustment power:

[0084]

[0085] Where, ΔP m Δf represents the regulation power required for the m-th air conditioning cluster, and Δf represents the power grid frequency deviation signal.

[0086] Furthermore, the specific steps of step five are as follows:

[0087] Step 5.1: Calculate the comprehensive score for each air conditioning terminal in the cluster sequentially. The calculation formula is as follows:

[0088] S i =ω1F1+ω2F2+ω3F3

[0089]

[0090] Among them, S i F1 represents the overall score of the i-th air conditioning terminal, F2 represents the temperature margin score, F3 represents the regulation capability score, and F4 represents the response speed score. ω1, ω2, and ω3 represent the average response time, and ω1, ω2, and ω3 represent the score weighting coefficients.

[0091] Step 5.2: Sort the air conditioning terminals by their comprehensive scores from highest to lowest, select them sequentially, and accumulate their adjustable power until the cumulative adjustment capacity is ≥ ΔP.m ;

[0092] Step 5.3: Distribute the adjustment amount obtained by each air conditioning terminal and execute it.

[0093] In this embodiment, in step 5.1, ω1 is 0.4, ω2 is 0.4, and ω3 is 0.2; the average response time is the average time from receiving the command to the start of power change.

[0094] In step 5.2, the adjustable power is the potential power adjustment range calculated based on the thermodynamic model and temperature margin.

[0095] Example 2

[0096] This invention proposes a virtual synchronous machine cluster collaborative control system for air conditioning loads, which applies the above-mentioned method for virtual synchronous machine cluster collaborative control of air conditioning loads, including:

[0097] The data acquisition module is used to acquire the operating parameters and power grid frequency deviation signals of multiple air conditioning terminals, and to preprocess the acquired data. Specifically, the operating parameters include indoor temperature, set temperature, power consumption, and external ambient temperature. The preprocessing includes operations such as outlier removal and missing value filling.

[0098] The candidate terminal screening module is used to screen adjustable candidate air conditioning terminals in the cluster based on the preprocessed operating parameters, exclude terminals that have just completed a start-up and shutdown, whose room temperature exceeds the comfortable temperature range, whose equipment is faulty or offline, and construct a multi-dimensional data vector for each candidate terminal.

[0099] The dynamic grouping module is used to evaluate the similarity between candidates based on their multidimensional data vectors by calculating the weighted normalized Euclidean distance between them, and dynamically divide the highly similar terminals into several time-varying clusters.

[0100] The virtual synchro modeling module is used to build a virtual synchro model for each time-varying cluster. It calculates the equivalent moment of inertia, equivalent damping coefficient, and droop gain of the model based on the heat capacity, number, and average regulation margin of the terminals in the cluster.

[0101] The regulation power calculation module is used to receive the grid frequency deviation signal and calculate the regulation power that each cluster should provide through a virtual synchronous machine model.

[0102] The instruction decomposition and distribution module is used to calculate the comprehensive score of each terminal in the cluster, select terminals according to the score and accumulate the adjustable power, decompose the adjustment instructions to the selected individual terminals and distribute them for execution; specifically, the comprehensive score combines temperature difference margin, adjustment capability and response speed.

[0103] The data storage and monitoring module is used to store the operating data, terminal status, and control results of each module, monitor the cluster's operating status and control effects in real time, and provide data support for subsequent optimization.

[0104] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A method for coordinated control of air conditioning load virtual synchronous machine clusters, characterized in that, Includes the following steps: Step 1: Collect operating parameters from multiple air conditioning terminals every 15 minutes, filter out adjustable candidate air conditioning terminals, and construct a multi-dimensional data vector for each candidate air conditioning terminal; Step 2: Use the multidimensional data vector to dynamically group the air conditioning terminals to form several time-varying clusters; Step 3: Construct a virtual synchronization machine model for each cluster; Step 4: Based on the power grid frequency deviation signal, calculate the regulation power that each cluster should provide using the virtual synchronous machine model; Step 5: Decompose the adjustment instructions to the individuals within the cluster and issue them for execution.

2. The method for coordinated control of air conditioning load virtual synchronous machine clusters according to claim 1, characterized in that: In step one, the operating parameters of the air conditioning terminal include the indoor temperature T. in Set temperature T set Power consumption P, building thermal time constant τ, ambient temperature T out ; The multidimensional data vector is: x i =[T in ,T set ,|T in -T set |,P,T out ,τ] Where, x i Let i represent the multidimensional data vector of the i-th air conditioner.

3. The method for coordinated control of air conditioning load virtual synchronous machine clusters according to claim 1, characterized in that: The specific steps of step two are as follows: Step 2.1: Normalize the multidimensional data vector; Step 2.2: Calculate the weighted normalized Euclidean distance between the air conditioning terminals sequentially. The calculation formula is as follows: Where, d ij ω represents the weighted normalized Euclidean distance between the i-th air conditioning terminal and the j-th air conditioning terminal. k σ represents the weight coefficient of the k-th dimension feature. k The standard deviation represents the k-th dimension feature; Step 2.3, set the distance threshold to θ, if d ij If ≤θ, it means that the i-th air conditioning terminal and the j-th air conditioning terminal have similar operating characteristics; otherwise, it means that they are not similar. Step 2.4: Count the number of air conditioning terminals with similar operating characteristics to each of the aforementioned air conditioning terminals, and sort them in descending order of quantity; Step 2.5: Group the top-ranked air conditioning terminal and all air conditioning terminals with similar operating characteristics into cluster G. m ; Step 2.6: Repeat steps 2.4 to 2.5 for the air conditioning terminals that are not assigned to a cluster, until all the air conditioning terminals are assigned to a cluster.

4. The method for coordinated control of air conditioning load virtual synchronous machine clusters according to claim 1, characterized in that: In step three, the calculation formula for the virtual synchronizer model parameters is as follows: K m =g·h m Among them, J m D represents the equivalent moment of inertia. m K represents the equivalent damping coefficient. m C represents droop gain or response sensitivity. i G represents the heat capacity of the i-th air conditioner. m Let |G| represent the m-th air conditioning cluster. m | represents the number of air conditioning terminals in the m-th air conditioning cluster, η m T represents the average regulation margin of the m-th air conditioning cluster, α, β, and γ represent coefficients. in,i Indicates the current indoor temperature, T low Indicates the minimum comfortable temperature, T high P represents the highest comfortable temperature. i P represents the current operating power. rated This indicates the rated maximum power.

5. The method for coordinated control of air conditioning load virtual synchronous machine clusters according to claim 1, characterized in that: In step four, each cluster should provide the following formula for calculating the adjustment power: Where, ΔP m Δf represents the regulation power required for the m-th air conditioning cluster, and Δf represents the power grid frequency deviation signal.

6. The method for coordinated control of air conditioning load virtual synchronous machine clusters according to claim 1, characterized in that: The specific steps of step five are as follows: Step 5.1: Calculate the comprehensive score for each air conditioning terminal in the cluster sequentially. The calculation formula is as follows: S i =ω1F1+ω2F2+ω3F3 Among them, S i F1 represents the overall score of the i-th air conditioning terminal, F2 represents the temperature margin score, F3 represents the regulation capability score, and F4 represents the response speed score. ω1, ω2, and ω3 represent the average response time, and ω1, ω2, and ω3 represent the score weighting coefficients. Step 5.2: Sort the air conditioning terminals by their comprehensive scores from highest to lowest, select them sequentially, and accumulate their adjustable power until the cumulative adjustment capacity is ≥ ΔP. m ; Step 5.3: Distribute the adjustment amount obtained by each air conditioning terminal and execute it.

7. A virtual synchronous machine cluster collaborative control system for air conditioning load, employing the virtual synchronous machine cluster collaborative control method for air conditioning load as proposed in any one of claims 1 to 7, characterized in that, include: The data acquisition module is used to acquire the operating parameters of multiple air conditioning terminals and the power grid frequency deviation signal, and to preprocess the acquired data. The candidate terminal screening module is used to screen adjustable candidate air conditioning terminals in the cluster based on the preprocessed operating parameters, exclude terminals that have just completed a start-up and shutdown, whose room temperature exceeds the comfortable temperature range, whose equipment is faulty or offline, and construct a multi-dimensional data vector for each candidate terminal. The dynamic grouping module is used to evaluate the similarity between candidates based on their multidimensional data vectors by calculating the weighted normalized Euclidean distance between them, and dynamically divide the highly similar terminals into several time-varying clusters. The virtual synchro modeling module is used to build a virtual synchro model for each time-varying cluster. It calculates the equivalent moment of inertia, equivalent damping coefficient, and droop gain of the model based on the heat capacity, number, and average regulation margin of the terminals in the cluster. The regulation power calculation module is used to receive the grid frequency deviation signal and calculate the regulation power that each cluster should provide through a virtual synchronous machine model. The instruction decomposition and distribution module is used to calculate the comprehensive score of each terminal in the cluster, select terminals according to the score and accumulate the adjustable power, decompose the adjustment instructions to the selected individual terminals and distribute them for execution; The data storage and monitoring module is used to store the operating data, terminal status, and control results of each module, monitor the cluster's operating status and control effects in real time, and provide data support for subsequent optimization.

Images