Air conditioner control method and system based on group control calculation
By adopting a group control computing-based air conditioning control method, dynamic grouping and load regulation of air conditioning equipment are realized, which solves the problems of slow and inaccurate air conditioning group control strategies in existing technologies, improves the adaptability and operating efficiency of air conditioning systems in complex environments, and realizes efficient collaborative control of air conditioning systems in complex environments.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- SATURN CHANGZHOU TECH
- Filing Date
- 2026-03-18
- Publication Date
- 2026-06-05
AI Technical Summary
Existing air conditioning group control technology is relatively static in its equipment grouping strategy, lacking dynamic perception and targeted analysis of local environmental disturbance factors within the area. This results in slow response and insufficient accuracy of the control strategy, and fails to effectively consider the periodic fluctuations in equipment load and changes in health status over time, affecting the operating efficiency and energy-saving effect of the air conditioning system.
A group control computing approach is adopted, which integrates equipment association grouping based on local thermal disturbance sensitivity, embeds control margin correction based on load fluctuation cycle characteristics and equipment health degradation status, and combines a negotiation allocation algorithm based on environmental thermal load disturbance prediction to generate load control commands for air conditioning equipment.
It improves the adaptability and response accuracy of the air conditioning system in complex environments, reduces local control conflicts, enhances regulation efficiency and stability, and strengthens the overall operating effect of the air conditioning system under complex environmental conditions.
Smart Images
Figure CN122149074A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of air conditioning control technology, and in particular to an air conditioning control method and system based on group control computing. Background Technology
[0002] In recent years, with the continuous improvement of building intelligence and refined management, intelligent operation technology for multi-device group control has developed rapidly. In practical engineering applications, air conditioning equipment often needs to be coordinated and controlled according to the complex and ever-changing environmental conditions in the area to ensure the uniformity and comfort of indoor temperature while reducing operating energy consumption. However, existing group control technologies are usually relatively static and extensive in terms of equipment grouping strategies, lacking dynamic perception and targeted analysis of local environmental disturbance factors in the area. This often results in slow response and insufficient accuracy of control strategies during actual operation, which seriously affects the overall operating efficiency and energy-saving effect of the air conditioning system.
[0003] Furthermore, the commonly used equipment load control margin is usually determined based on the rated operating range of the equipment, without taking into account the characteristics of the periodic fluctuation of equipment load over time and the dynamic changes in the health status of the equipment during actual operation. This may lead to a mismatch between the margin calculation and the actual capacity of the equipment during actual operation, further reducing the reliability and economy of the air conditioning group control strategy.
[0004] Therefore, how to achieve accurate prediction of environmental disturbances by air conditioning group control strategies and fully combine the actual operating characteristics and status of equipment for dynamic and intelligent collaborative control has become an important technical problem that urgently needs to be solved in this field. Summary of the Invention
[0005] The present invention aims to solve at least one of the technical problems existing in the prior art; to this end, the present invention proposes an air conditioning control method and system based on group control computing.
[0006] To achieve the above objectives, the present invention provides the following technical solution: In a first aspect, the present invention provides an air conditioning control method based on group control computing, comprising: Based on the real-time operating load data of each air conditioning unit and the real-time monitoring temperature data of the target area, a clustering analysis method that integrates local thermal disturbance sensitivity is used to generate equipment association groups; Based on the historical operating load data of each air conditioning unit, the periodic fluctuation pattern of the load is extracted to generate the equipment control margin embedded with the periodic characteristics of the load fluctuation. A regression fitting method incorporating degradation rate perception was used to analyze the maintenance records and cumulative operating time of each air conditioning unit, generating load margin correction coefficients that characterize the health degradation level of the equipment. By using a negotiation allocation algorithm that incorporates environmental heat load disturbance prediction, the associated grouping of equipment, equipment control margin, and load margin correction coefficient are calculated collaboratively to generate load control commands for air conditioning equipment.
[0007] Further, the generating device associated group includes: Based on a preset spatial grid, the real-time temperature data of the target area is discretized to obtain spatial hotspot regions that characterize local thermal disturbance events; A sensitivity analysis method with embedded hotspot area spatial influence attenuation factor is used to construct the sensitivity intensity relationship between air conditioning equipment and spatial hotspot areas; Based on the sensitivity intensity relationship, a clustering method with enhanced local sensitivity is used to group the air conditioning equipment, resulting in equipment association groups.
[0008] Furthermore, the construction of the sensitivity intensity relationship between the air conditioning equipment and the spatial hotspot area includes: Based on the physical distance between the equipment and the hotspot area and the nonlinear attenuation characteristics of thermal disturbance propagation, a spatial influence attenuation factor is constructed. By combining the spatial influence attenuation factor with the historical response characteristics of the equipment, a sensitivity intensity relationship reflecting the influence of spatial location is generated.
[0009] Furthermore, the device control margin for generating embedded load fluctuation cycle characteristics includes: Extract the periodic peak-valley time difference characteristics of the load changes of each air conditioning unit to determine the load fluctuation period of the unit; The load fluctuation cycle is verified using a load cycle stability verification method to form a verified load fluctuation cycle. Based on the verified load fluctuation cycle and the rated operating regulation capacity of the equipment, a control margin for the equipment embedded with the load fluctuation cycle characteristics is generated.
[0010] Furthermore, the method for verifying load fluctuation cycles using a load cycle stability verification method includes: Based on historical load fluctuation data from multiple cycles, the differences in the timing of peaks and troughs in each cycle are calculated. A robustness assessment method based on the distribution of peak and valley time differences was adopted to confirm the stability level of each cycle and to sort and screen the cycle stability. Typical load fluctuation cycles are determined from the stability cycles selected through sorting and screening, and used as the verified load fluctuation cycles.
[0011] Furthermore, the load margin correction coefficient for generating the characterization of the equipment health degradation level includes: Extract equipment load response change data and cumulative equipment runtime from the maintenance records of each equipment to construct the equipment load response degradation trajectory. A nonlinear regression method that integrates the preset equipment lifespan expectation and the equipment load response degradation trajectory is used to fit the health degradation rate of the equipment; Equipment health degradation levels are classified according to the rate of health deterioration, and load margin correction coefficients are generated.
[0012] Furthermore, the nonlinear regression method that integrates preset equipment lifespan expectations and equipment load response degradation trajectories includes: Based on the equipment load response degradation trajectory, a nonlinear degradation fitting function considering the preset equipment life expectation constraint is constructed; A nonlinear fitting method based on least squares support vector machine is used to solve the nonlinear degradation fitting function and obtain the health degradation rate of the equipment.
[0013] Furthermore, the generation of load control instructions for the air conditioning equipment includes: Based on historical environmental disturbance data and regional load change data, a recursive prediction model with embedded time-series memory mechanism is used to predict environmental heat load disturbances. Based on the predicted environmental heat load disturbance results and equipment association grouping, a negotiation game algorithm is used to determine the initial load compensation requirements of each equipment group; A margin optimization allocation algorithm combining equipment control margin and load margin correction coefficient is adopted to perform collaborative optimization calculation on the initial load compensation demand and generate load control commands for air conditioning equipment.
[0014] Furthermore, the method of determining the initial load compensation requirements for each equipment group using a negotiation game algorithm includes: A negotiation game framework with environmental heat load disturbance as the external constraint is constructed, and the negotiation payoff function of equipment grouping is defined. Based on the negotiated payoff function among the equipment groups under the game framework, the load compensation demand solution that satisfies the global disturbance demand balance is determined, and the initial load compensation demand of the equipment groups is generated.
[0015] Secondly, the present invention provides an air conditioning control system based on group control computing, implemented based on the aforementioned air conditioning control method based on group control computing, the system comprising: Sensitive Grouping Module: This module generates equipment association groups based on the real-time operating load data of each air conditioning unit and the real-time monitored temperature data of the target area, using a clustering analysis method that incorporates the sensitivity to local thermal disturbances. Cycle margin module: used to extract the periodic fluctuation pattern of the load based on the historical operating load data of each air conditioning unit, so as to generate the equipment control margin embedded with the load fluctuation cycle characteristics; Deterioration correction module: Used to analyze the maintenance records and cumulative operating time of each air conditioning unit using a regression fitting method that integrates degradation rate perception, and generate load margin correction coefficients that characterize the health degradation level of the equipment; Negotiation and Allocation Module: This module uses a negotiation and allocation algorithm embedded with environmental heat load disturbance prediction to collaboratively calculate equipment association grouping, equipment control margin, and load margin correction coefficient, thereby generating load control commands for air conditioning equipment.
[0016] Compared with the prior art, the beneficial effects of the present invention are: This invention introduces a device association grouping method that integrates the sensitivity of local environmental disturbances. This method enables the grouping of air conditioning devices to no longer rely on static preset divisions, but to be dynamically divided based on the spatial influence intensity of temperature disturbances in the area. This improves the adaptability and response accuracy of the group control strategy in changing environments, enhances the collaborative performance of the air conditioning system, and effectively reduces local control conflicts.
[0017] This invention establishes control margins and correction coefficients based on the historical load fluctuation characteristics and health degradation status of equipment, enabling each air conditioning unit to have a quantifiable actual control capability evaluation during load adjustment. This avoids problems of insufficient or excessive control caused by a single static margin setting, thereby improving the control efficiency of the overall group control strategy and the stability of equipment operation.
[0018] This invention employs a load control strategy that combines environmental heat load disturbance prediction with negotiated allocation, enabling group control scheduling to respond in advance to future load change trends. Under the constraints of equipment association grouping and control margin, it achieves optimized calculation of load allocation, improving the foresight, coordination, and energy efficiency of group control scheduling, and enhancing the overall operating effect of the air conditioning system under complex environmental conditions. Attached Figure Description
[0019] Figure 1 This is a flowchart of an air conditioning control method based on group control computing, as shown in Example 1.
[0020] Figure 2 This is an architecture diagram of an air conditioning control system based on group control computing, as shown in Example 2. Detailed Implementation
[0021] The technical solution of the present invention will be clearly and completely described below with reference to the embodiments. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.
[0022] Example 1: Please see Figure 1This invention provides an air conditioning control method based on group control computing, comprising: Based on the real-time operating load data of each air conditioning unit and the real-time monitoring temperature data of the target area, a clustering analysis method that integrates local thermal disturbance sensitivity is used to generate equipment association groups; In a specific implementation, the generation device association group includes: Based on a preset spatial grid, the real-time temperature data of the target area is discretized to obtain spatial hotspot regions that characterize local thermal disturbance events; In one specific embodiment, in order to effectively capture the non-uniform distribution characteristics of local temperature in the area, a standardized spatial grid is first constructed in the area to be monitored; for example, the indoor space is divided into multiple spatial grid units at equal intervals, and each grid unit corresponds to a unique spatial location coordinate.
[0023] Then, based on the real-time temperature monitoring data of each grid cell, the real-time temperature value of each grid cell is calculated.
[0024] Furthermore, the grid cells are density-clustered using spatial hotspot identification methods (such as spatial clustering based on coordinate location or heat map analysis methods) to determine the set of spatial grid cells with temperatures greater than or equal to a set threshold. The geometric center or weighted centroid of the set is marked as the spatial hotspot region of the local thermal disturbance event, thereby providing a coordinate reference for subsequent calculation of sensitivity intensity.
[0025] For example, if the average temperature of all grid cells in the area to be monitored is 25°C at a certain moment, and the temperature threshold is set to be 2°C above the average temperature, then grid cells with a real-time temperature greater than or equal to 27°C are defined as hotspot areas, thereby achieving accurate identification of spatial hotspot areas.
[0026] A sensitivity analysis method with embedded hotspot area spatial influence attenuation factor is used to construct the sensitivity intensity relationship between air conditioning equipment and spatial hotspot areas; It should be understood that, in order to accurately reflect the actual impact of temperature disturbances in hotspot areas on different air conditioning equipment, this embodiment introduces a hotspot area spatial impact attenuation factor to quantitatively characterize the intensity of the spatial impact of hotspot areas on surrounding air conditioning equipment.
[0027] In a specific implementation, the construction of the sensitivity intensity relationship between the air conditioning equipment and the spatial hotspot area includes: Based on the physical distance between the equipment and the hotspot area and the nonlinear attenuation characteristics of thermal disturbance propagation, a spatial influence attenuation factor is constructed. Specifically, the spatial influence attenuation factor is calculated using a nonlinear distance attenuation function, and the calculation formula is as follows: In the formula, This represents the spatial impact attenuation factor between the i-th air conditioning unit and the j-th hotspot area. This represents the actual spatial distance between the i-th air conditioning unit and the j-th hotspot area. Indicates the attenuation reference distance. This represents the decay index.
[0028] It should be noted that the spatial influence attenuation factor The coordinate mapping relationship between each device and the hotspot area in the indoor layout diagram is calculated. Specifically, a two-dimensional Euclidean coordinate system is established with a preset reference point in the target area as the origin. Based on the preset coordinates of each device in the layout diagram and the real-time coordinates of the geometric center of the hotspot area, the coordinates are calculated using the Euclidean distance formula. Attenuation reference distance and decay index Based on experimental results, in this embodiment, the attenuation reference distance is... The value is 5 meters, and the attenuation index is... The value range is 1.5 to 2.5. For example, if the first air conditioner is 3 meters away from the hotspot area and the second air conditioner is 6 meters away from the hotspot area, the value is set... rice, Then they can be calculated separately: Attenuation factor of the first air conditioning unit Attenuation factor of the second air conditioning unit It is understood that the larger the spatial impact attenuation factor, the greater the spatial impact of the hotspot area on the corresponding air conditioning equipment.
[0029] By combining the spatial influence attenuation factor with the historical response characteristics of the equipment, a sensitivity intensity relationship reflecting the influence of spatial location is generated.
[0030] Specifically, to accurately characterize the differences in sensitivity response of different air conditioning devices to temperature disturbance events in hotspot areas, this embodiment analyzes the actual response records of the devices during historical operation to similar hotspot disturbance events. The specific calculation method is as follows: First, extract the actual response amplitude of the equipment when temperature events occur in hot spots from historical operating data, such as the adjustment amplitude of the equipment's air outlet temperature and the change amplitude of the air supply volume. Secondly, calculate the historical average response amplitude of each air conditioning unit to the hotspot area, and record it as the historical response coefficient. Finally, the historical response coefficients By weighting and fusing the calculated spatial influence attenuation factor, the sensitivity intensity relationship is obtained: In the formula, This represents the sensitivity intensity relationship value of the i-th air conditioning unit to the j-th hotspot area. For example, if the historical response coefficient of the first air conditioning unit is 0.6 and that of the second air conditioning unit is 0.8, and combined with the above attenuation factor calculation results, the sensitivity intensity relationship of the first air conditioning unit is 0.4193 and that of the second unit is 0.1895.
[0031] Based on the sensitivity intensity relationship, a clustering method with enhanced local sensitivity is used to group the air conditioning equipment, resulting in equipment association groups.
[0032] In a specific embodiment, to ensure that the generated device groups effectively reflect actual spatial sensitivity, this embodiment selects a clustering method that strengthens sensitivity relationships, specifically: First, based on the calculated sensitivity intensity relationship matrix, the sensitivity features of the air conditioning equipment are represented as high-dimensional feature vectors. Each element of the feature vector is the sensitivity intensity value of the corresponding equipment to each hot spot area. When the number of hot spots is different at different times, zero padding or dimensionality reduction is used to keep the feature vector dimension consistent. Secondly, a clustering algorithm based on sensitivity differences is selected; specifically, this embodiment uses a sensitivity-weighted k-means clustering algorithm to calculate the sensitivity intensity of each device to different hotspot areas. As the dimension weights of the feature vectors, when calculating the Euclidean distance, higher weight coefficients are assigned to highly sensitive dimensions, thereby enhancing the discriminative power of spatial thermal perturbation sensitivity differences in the clustering process. Finally, the device association groups are generated through iterative optimization using a clustering algorithm.
[0033] For example, if there are 20 air conditioning units and 3 hotspot areas, the sensitivity intensity matrix is 20×3-dimensional data; after optimization and iteration using the above clustering method, the following specific grouping results can be obtained: Equipment Group 1: Air conditioning equipment {1,3,5,7,8}, which is highly sensitive to hotspot area 1; Equipment group 2: Air conditioning equipment {2,4,6,9,10,11}, which is highly sensitive to hotspot area 2; Equipment group 3: Air conditioning equipment {12,13,14,15,16,17,18,19,20}, which is highly sensitive to hotspot area 3.
[0034] It should be noted that the device numbers and grouping in the above examples are merely examples, and those skilled in the art can adjust them according to the actual device layout, sensitivity characteristics, and algorithm parameters to obtain specific grouping information.
[0035] Understandably, the device association groups generated by the above method can accurately reflect the actual sensitivity of air conditioning equipment to spatial hotspot disturbance events, ensuring the accuracy and effectiveness of the device collaborative control strategy in subsequent group control calculations.
[0036] Based on the historical operating load data of each air conditioning unit, the periodic fluctuation pattern of the load is extracted to generate the equipment control margin embedded with the periodic characteristics of the load fluctuation. In a specific implementation, the equipment control margin for generating embedded load fluctuation cycle characteristics includes: Extract the periodic peak-valley time difference characteristics of the load changes of each air conditioning unit to determine the load fluctuation period of the unit; It should be noted that this embodiment first obtains the time series information of the historical operating load data of the air conditioning equipment. The historical operating load data specifically includes the load values recorded in real time during the long-term operation of each air conditioning equipment. The load values can be parameters that can directly reflect the operating load status of the equipment, such as the real-time operating power, cooling capacity, or current of the equipment.
[0037] Furthermore, this embodiment utilizes time series analysis methods to extract the periodic characteristics of load data; for example, autocorrelation analysis or Fourier spectrum analysis methods are used to obtain the periodic characteristics of load fluctuations.
[0038] For example, this embodiment uses the autocorrelation analysis method for calculation, and the specific formula is as follows: In the formula, This represents the value of the autocorrelation function. This represents the load value of the air conditioning equipment at time point t. This represents the historical average of the load value. This represents the time delay step, and N represents the total number of historical data sample points.
[0039] By analyzing the peak position of the autocorrelation function, the period length of the load change of the air conditioning equipment can be accurately determined. For example, in the analysis of specific historical data, if the first peak of the autocorrelation function appears with a time delay of 24 hours, it indicates that the equipment load fluctuation period is 24 hours.
[0040] The load fluctuation cycle is verified using a load cycle stability verification method to form a verified load fluctuation cycle. In one specific embodiment, in order to ensure the reliability and stability of the load cycle characteristics of the air conditioning equipment, this embodiment further introduces a cycle stability verification method to effectively verify the cycle characteristics.
[0041] In implementation, the method for verifying load fluctuation cycles using a load cycle stability verification method includes: Based on historical load fluctuation data from multiple cycles, the differences in the timing of peaks and troughs in each cycle are calculated. Specifically, this embodiment first divides the historical data into periods to determine several complete load data intervals, and then identifies the peak and valley times of the air conditioning equipment load data in each period interval.
[0042] For example, if the load data sampling interval is set to 10 minutes and one cycle is 24 hours, then each cycle contains 144 sampling points. Subsequently, in this embodiment, the specific times of the peak and valley values in each cycle are marked, and the difference between the peak and valley values between consecutive cycles is calculated. For example, if the peak time of the first cycle is 14:00 and the peak time of the second cycle is 14:20, then the difference between the peak times is 20 minutes. The difference between the peak and valley times between all cycles is calculated in this way.
[0043] A robustness assessment method based on the distribution of peak and valley time differences was adopted to confirm the stability level of each cycle and to sort and screen the cycle stability. In this implementation, robust statistical methods (such as the coefficient of variation, CV) are used for stability assessment. The specific calculation formula is as follows: Where CV represents the coefficient of variation. The standard deviation of the peak-valley time difference data. This represents the average value of the difference between peak and valley times.
[0044] The stability of the load cycle is assessed by calculating the coefficient of variation. The smaller the coefficient of variation, the higher the periodic stability.
[0045] Furthermore, the stability results of each period are sorted according to the size of the coefficient of variation, and a threshold for the coefficient of variation (e.g., 0.1) is set. Only periods with a coefficient of variation less than the threshold are retained to ensure that the finally selected periods have sufficient stability and representativeness.
[0046] Typical load fluctuation cycles are determined from the stability cycles selected through sorting and screening, and used as the verified load fluctuation cycles.
[0047] It should be noted that in this embodiment, a typical cycle is selected from the cycles after stability screening as the standard load fluctuation cycle of the equipment. The specific selection method is as follows: First, perform load waveform similarity assessment on cycles with acceptable stability (such as cosine similarity calculation or dynamic time warping (DTW) distance calculation) to calculate the similarity or distance between the load fluctuation of each cycle and the historical average load fluctuation. Then, the cycle with the highest similarity (or smallest distance) is defined as the typical load fluctuation cycle, which is used for subsequent calculation of the control margin of air conditioning equipment.
[0048] It is understandable that the typical load fluctuation cycle obtained in the above manner can accurately reflect the typical periodic fluctuation characteristics of the equipment operating load, providing sufficient data support for the accurate calculation of equipment control margin in subsequent steps.
[0049] Based on the verified load fluctuation cycle and the rated operating regulation capacity of the equipment, a control margin for the equipment embedded with the load fluctuation cycle characteristics is generated.
[0050] Specifically, in order to reasonably quantify the ability of each air conditioning unit to dynamically regulate load cycle fluctuations during actual operation, this embodiment introduces the quantitative indicator of equipment regulation margin to represent the range of air conditioning units' ability to respond to actual regulation needs under periodic load fluctuation modes.
[0051] In a specific embodiment, the method for calculating the equipment control margin is as follows: First, based on the verified typical load fluctuation cycle, determine the maximum peak value and minimum valley value of the equipment operating load within the cycle, and use this as the periodic load benchmark for the air conditioning equipment; for example, in the load fluctuation curve determined by the aforementioned typical load cycle data, the maximum peak load is 80% of the rated power, and the minimum valley load is 30% of the rated power.
[0052] Secondly, based on the rated operating adjustment capability of the air conditioning equipment, determine the maximum load control range that the air conditioning equipment can actually achieve. Specifically, the rated operating adjustment capability includes, but is not limited to, the maximum adjustable range of the rated power of the air conditioning equipment, the power increase / decrease rate, the response delay time, and other technical parameters. For example, the rated operating adjustment capability of a certain model of air conditioning equipment is a maximum adjustment range of ±25% of the rated power, a load increase / decrease rate of 5% per minute, and a load response delay of less than 1 minute.
[0053] Furthermore, by combining the maximum peak and minimum trough values within a typical load fluctuation cycle with the rated operating regulation capacity of the air conditioning equipment, the specific value of the equipment's regulation margin can be calculated and expressed as: In the formula, This represents the control margin value of the i-th air conditioning unit. This indicates the maximum power adjustment limit that the air conditioning equipment can achieve (e.g., 125% of rated power). This indicates the lower limit of the power adjustment for air conditioning equipment (e.g., 20% of rated power). These represent the peak and trough values of the load during a typical load fluctuation cycle, respectively. Indicates the rated power of the air conditioning equipment. This indicates the maximum rated margin specified for the air conditioning equipment (e.g., 25% of the rated power).
[0054] For example, suppose an air conditioning unit has a rated power of 10kW and a peak load during a typical load fluctuation period. 8kW (i.e., 80% of rated capacity), valley load The maximum power regulation limit for the equipment is 3kW (i.e., 30% of rated power). The minimum power regulation limit for stable operation of the equipment is 12.5kW (i.e., 125% of rated power). The maximum rated capacity of the equipment is 2kW (i.e., 20% of the rated capacity). The margin is 25%. Based on the formula, the uplink margin is (12.5-8) / 10=45%, and the downlink margin is (3-2) / 10=10%. In this embodiment, to ensure the symmetry and stability of the equipment control, the final control margin is... Take the minimum value among the upside margin, downside margin, and nominal margin, i.e. .
[0055] It should be understood that if the calculated downside margin is negative at a certain moment (e.g., the valley load is lower than the stable operating limit), it indicates that the equipment does not have the ability to adjust downside in that cycle. Through the above calculation, it is ensured that the generated control margin is always within the intersection of the equipment's physical performance and the periodic load benchmark.
[0056] A regression fitting method incorporating degradation rate perception was used to analyze the maintenance records and cumulative operating time of each air conditioning unit, generating load margin correction coefficients that characterize the health degradation level of the equipment. In a specific implementation, the load margin correction coefficient for generating the equipment health degradation level includes: Extract equipment load response change data from maintenance nodes in the equipment maintenance records; In this embodiment, in order to objectively reflect the changing trend of the health status of the air conditioning equipment, the relevant maintenance record data of each air conditioning equipment at each maintenance node are first collected; the data of the maintenance node is specifically the actual performance measurement data of the equipment before the maintenance action is carried out, so as to truly record the natural deterioration characteristics of the equipment over time; the maintenance record data focuses on the changes in the load response characteristics of the equipment, specifically including the difference or rate of change of key performance indicators such as the output cooling power or supply air temperature difference of the equipment before and after the maintenance node.
[0057] For example, if the load response characteristics (e.g., cooling capacity) of an air conditioning unit before a certain maintenance point are 100% of the rated capacity, and the cooling capacity after the maintenance point is measured to be 92% of the rated capacity, then it can be determined that the load response change data at that maintenance point is a decrease of 8%.
[0058] By combining the cumulative operating time of the equipment and the data on changes in the equipment load response, a degradation trajectory of the equipment load response is constructed; Specifically, in order to more fully and accurately describe the entire process of performance degradation during the operation of air conditioning equipment, this embodiment combines the cumulative runtime information of the equipment to construct the equipment load response degradation trajectory; the cumulative runtime is specifically the number of operating hours accumulated by the air conditioning equipment from initial operation to maintenance node.
[0059] Furthermore, the load response change data of each maintenance node is mapped one-to-one with the cumulative runtime of the equipment to form discrete equipment load response degradation data points, thereby obtaining historical sequence data that can accurately reflect the trend of equipment performance degradation.
[0060] For example, the actual maintenance record data of a certain air conditioning equipment is shown in Table 1 below.
[0061]
[0062] The above data can be used to obtain the equipment load response degradation trajectory, reflecting the gradual performance degradation trend of air conditioning equipment during operation.
[0063] A nonlinear regression method that integrates the preset equipment lifespan expectation and the equipment load response degradation trajectory is used to fit the health degradation rate of the equipment; Specifically, this embodiment uses a nonlinear regression method to fit the equipment load response degradation trajectory to obtain a mathematical expression that accurately describes the rate of degradation of the equipment's health status; this embodiment introduces the preset lifespan expectation of the air conditioning equipment as a constraint condition to ensure the rationality of the regression fitting.
[0064] In a specific implementation, the nonlinear regression method that integrates preset equipment lifespan expectations and equipment load response degradation trajectories includes: Based on the equipment load response degradation trajectory, a nonlinear degradation fitting function considering the preset equipment life expectation constraint is constructed; It should be noted that the nonlinear degradation fitting function can be in exponential or power function form; this embodiment uses an exponential degradation fitting function: In the formula, This represents the load response characteristic value of the air conditioning equipment when the cumulative running time is t. The parameters a, b, and c represent regression coefficients, where b is a negative value used to characterize the nonlinear decay rate of equipment performance over time, a is the decay amplitude coefficient, which, together with the constant term c, determines the initial performance level of the equipment at t=0, and c characterizes the asymptotic limit value of the equipment's performance after long-term operation; t is the cumulative running time of the air conditioning equipment.
[0065] The parameter constraint is that when the equipment reaches the preset end of its lifespan (e.g., 8000 hours), The value should not be less than the lower limit of the design performance of the air conditioning equipment (e.g., 70% of the rated performance).
[0066] It should be noted that the preset lifespan end and design performance lower limit are specifically determined by the equipment's technical parameters.
[0067] A nonlinear fitting method based on least squares support vector machine is used to solve the nonlinear degradation fitting function and obtain the health degradation rate of the equipment.
[0068] Specifically, in this embodiment, the least squares support vector machine regression method is used to optimize the parameters of the above nonlinear fitting function in order to accurately obtain the health degradation rate of the equipment.
[0069] The implementation process of the least squares support vector machine regression method is as follows: First, using the cumulative operating time t of the equipment as the input variable, and the load response characteristic value... As an output variable, training sample data is extracted from the equipment load response degradation trajectory to establish a training sample set; Secondly, an LS-SVM kernel function based on radial basis function (RBF) is constructed, and the regularization parameters and kernel parameters of LS-SVM are determined by cross-validation or grid search. The regression model parameters of LS-SVM are optimized by minimizing the sum of squared fitting errors as the objective function.
[0070] By solving the LS-SVM model, a degradation curve that can characterize the continuous change of air conditioning equipment performance over time is obtained. The rate of health degradation of air conditioning equipment can be determined by calculating the slope or prediction deviation of the fitted curve at the current time point (or a specified time interval).
[0071] Equipment health degradation levels are classified according to the rate of health deterioration, and load margin correction coefficients are generated.
[0072] Specifically, in this embodiment, based on the health degradation rate obtained from the above fitting, different health degradation levels are defined and corresponding margin correction coefficients are given. The specific implementation process is as follows: First, a fitting model is used to predict the load response characteristics of the air conditioning equipment under the current cumulative operating time. The relative deviation between the current actual measured value and the predicted value is calculated, and the actual degree of deterioration of the air conditioning equipment is quantified by the relative deviation. The calculation formula is as follows: In the formula, This represents the health degradation deviation value of the i-th air conditioning unit. This represents the currently measured load response characteristic value of the air conditioning equipment. This represents the characteristic value of the fitted predicted load response.
[0073] Secondly, the equipment health status is divided into multiple degradation levels based on the magnitude of the deviation value; in this embodiment, the following level division rules are set: Level I (Excellent): When At that time, the equipment performance was in a relatively good state; Level II (Good): when At that time, the equipment was in good working order; Level III (Medium): When At that time, the equipment performance was at a moderate level of degradation; Grade IV (Poor): When At that time, the equipment performance was severely degraded.
[0074] Then, the load margin correction factor corresponding to the equipment is determined based on the degradation level; in this embodiment, the specific rules for setting the correction factor are as follows: The correction factor for Class I air conditioning equipment is 1.0; The correction factor for Class II air conditioning equipment is 0.9; The correction factor for Class III air conditioning equipment is 0.8; The correction factor for Class IV air conditioning equipment is 0.6.
[0075] It is understandable that the equipment load margin correction coefficient generated in the above manner reflects the true state of each air conditioning device under the influence of health degradation factors during actual operation, and can provide equipment status basis for subsequent air conditioning group control calculations.
[0076] By using a negotiation allocation algorithm that incorporates environmental heat load disturbance prediction, the associated grouping of equipment, equipment control margin, and load margin correction coefficient are calculated collaboratively to generate load control commands for air conditioning equipment.
[0077] In a specific implementation, the generation of load control instructions for the air conditioning equipment includes: Based on historical environmental disturbance data and regional load change data, a recursive prediction model with embedded time-series memory mechanism is used to predict environmental heat load disturbances. Specifically, this embodiment first collects historical disturbance data of the environment in which the air conditioning system is located, including environmental factors such as outdoor temperature, humidity, and solar radiation intensity, as well as historical load change data of the region. The above data is preprocessed to construct a time series data training set for training a recursive prediction model with a time series memory mechanism, such as a deep learning prediction model based on a long short-term memory network.
[0078] In practice, the input of the prediction model is a sequence of historical environmental data within a certain time window (e.g., the past 24 hours), and the output is a predicted value of environmental heat load disturbance within a specified future time period (e.g., the next hour).
[0079] Furthermore, during the prediction process, the currently collected real-time environmental data is input into the trained prediction model. Through the historical disturbance patterns learned within the model, the predicted environmental heat load disturbance results for the near future are obtained. For example, if the prediction model outputs that the ambient temperature will rise by 2°C within the next hour, the overall heat load of the corresponding area is expected to increase by approximately 10%. Based on the predicted environmental heat load disturbance results and the associated equipment groups, a negotiation game algorithm is used to determine the initial load compensation requirements for each equipment group. It is understood that this embodiment treats each associated equipment group as a participant in a negotiation game, and the environmental heat load disturbance prediction result as the overall load demand constraint outside the game. In specific implementation, the total predicted value of environmental heat load disturbance is first initially divided according to the area ratio of the area covered by each equipment group or the preset heat load weight, so as to obtain the ideal load compensation demand of each equipment group. Subsequently, each equipment group, through a negotiation game algorithm, combined its actual control capabilities with its ideal load compensation requirements. The process is iterated to determine the final initial load compensation requirement.
[0080] In a specific implementation, the step of using a negotiation game algorithm to determine the initial load compensation requirements for each equipment group includes: A negotiation game framework with environmental heat load disturbance as the external constraint is constructed, and the negotiation payoff function of equipment grouping is defined. It should be noted that the negotiated revenue function for equipment groups comprehensively considers the difference between the load compensation demand and the actual load compensation amount undertaken by each group, the overall controllability margin of the equipment within the group, and the constraints of equipment health status.
[0081] The revenue function for equipment grouping is specifically defined as the weighted sum of load compensation difference costs and equipment margin constraint costs. A higher revenue function value indicates a better load allocation strategy. The revenue function can be expressed as: In the formula, Let g represent the revenue function of the g-th device group. This represents the actual load compensation amount undertaken by the g-th equipment group. The prediction results indicate the ideal load compensation demand for the g-th device group. These represent the weight parameters of load variance and margin constraint in the revenue function, respectively. This represents the overall margin constraint value for the g-th device group.
[0082] It should be noted that the overall margin constraint for grouping The control margin of each device within this group and load margin correction factor The decision is made jointly, and the calculation method is as follows: In the formula, n represents the total number of devices in the g-th group. It should be understood that the overall margin constraint of the grouping This comprehensively reflects the actual adjustable depth of the group under the current health state; in the payoff function, by assigning... Positive correlation coefficient This allows groups with higher control margins and better health to have higher potential for gains in game negotiation, thus giving them priority in undertaking load compensation tasks that match their actual capabilities.
[0083] Based on the negotiated payoff function among the equipment groups under the game framework, the load compensation demand solution that satisfies the global disturbance demand balance is determined, and the initial load compensation demand of the equipment groups is generated.
[0084] In practice, the iterative negotiation process between each equipment group is realized through a negotiation game algorithm. Each equipment group dynamically adjusts its own load compensation demand with the goal of optimizing its own negotiation benefit function.
[0085] During each round of negotiation, each group updates its own load compensation decision based on the current decision status of other groups until the load compensation demand of each group stabilizes and converges to an equilibrium state, and the sum of the load compensation demand of all groups meets the predicted environmental heat load disturbance demand, and finally determines the initial load compensation demand of each equipment group.
[0086] For example, if the initial predicted increase in total load due to environmental disturbance is 100kW, after running the game negotiation algorithm, the load compensation demand for the first group is determined to be 40kW, the second group is 35kW, and the third group is 25kW. Finally, the sum of the load compensation demands of the three groups exactly meets the environmental disturbance demand.
[0087] A margin optimization allocation algorithm combining equipment control margin and load margin correction coefficient is adopted to perform collaborative optimization calculation on the initial load compensation demand and generate load control commands for air conditioning equipment.
[0088] In specific implementation, this embodiment further employs optimization calculation methods (such as linear programming or convex optimization) to allocate the initial load compensation requirements of the equipment group to specific air conditioning equipment, so as to ensure that the load control instructions allocated to each equipment are within the actual margin range.
[0089] The optimization objective is to minimize the sum of the differences between the load commands and the initial allocation requirements for each device, with the following constraints: The load control command for each device must be greater than or equal to 0, and less than or equal to the actual adjustable power limit of the device (i.e., rated power of the device × control margin of the device × load margin correction coefficient); the sum of the load compensation requirements allocated to each device is equal to the initial load compensation requirements determined by the group negotiation.
[0090] After optimization calculations, the final load control commands for each device can be determined.
[0091] For example, the control margin of a single air conditioning unit within a certain group Given a rated power of 10kW and a load margin correction factor of 0.9 (in a second-level healthy state), the upper limit of the final load control command for this equipment is: The calculation results ensure that the control commands do not exceed the physical adjustment limits of the equipment, and also avoid the operational risks caused by equipment aging and periodic load peaks.
[0092] Finally, the generated equipment load control commands are sent to the air conditioning equipment controller, and the equipment implements corresponding load control actions according to the commands, thereby realizing precise and coordinated control of the air conditioning system in the actual operating environment.
[0093] The air conditioning load control commands obtained by the above method fully consider the prediction results of environmental heat load disturbance, the spatial correlation of equipment, the periodic load fluctuation margin of equipment and the health degradation status, which can effectively improve the overall operational stability, energy saving and reliability of the air conditioning system.
[0094] Example 2:
[0095] like Figure 2 As shown, this embodiment discloses an air conditioning control system based on group control computing. For details not covered in this embodiment, please refer to the relevant parts of the description in Embodiment 1 above. The system includes: Sensitive Grouping Module: This module generates equipment association groups based on the real-time operating load data of each air conditioning unit and the real-time monitored temperature data of the target area, using a clustering analysis method that incorporates the sensitivity to local thermal disturbances. Cycle margin module: used to extract the periodic fluctuation pattern of the load based on the historical operating load data of each air conditioning unit, so as to generate the equipment control margin embedded with the load fluctuation cycle characteristics; Deterioration correction module: Used to analyze the maintenance records and cumulative operating time of each air conditioning unit using a regression fitting method that integrates degradation rate perception, and generate load margin correction coefficients that characterize the health degradation level of the equipment; Negotiation and Allocation Module: This module uses a negotiation and allocation algorithm embedded with environmental heat load disturbance prediction to collaboratively calculate equipment association grouping, equipment control margin, and load margin correction coefficient, thereby generating load control commands for air conditioning equipment.
[0096] Some of the data in the above formula are calculated by removing dimensions and taking their numerical values. The formula is the closest to the real situation obtained by software simulation of a large amount of collected data. The preset parameters and preset thresholds in the formula are set by those skilled in the art according to the actual situation or obtained through simulation of a large amount of data.
[0097] The above embodiments are only used to illustrate the technical methods of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical methods of the present invention without departing from the spirit and scope of the technical methods of the present invention.
Claims
1. An air conditioning control method based on group control computing, characterized in that, include: Based on the real-time operating load data of each air conditioning unit and the real-time monitoring temperature data of the target area, a clustering analysis method that integrates local thermal disturbance sensitivity is used to generate equipment association groups; Based on the historical operating load data of each air conditioning unit, the periodic fluctuation pattern of the load is extracted to generate the equipment control margin embedded with the periodic characteristics of the load fluctuation. A regression fitting method incorporating degradation rate perception was used to analyze the maintenance records and cumulative operating time of each air conditioning unit, generating load margin correction coefficients that characterize the health degradation level of the equipment. By using a negotiation allocation algorithm that incorporates environmental heat load disturbance prediction, the associated grouping of equipment, equipment control margin, and load margin correction coefficient are calculated collaboratively to generate load control commands for air conditioning equipment.
2. The air conditioning control method based on group control computing according to claim 1, characterized in that, The generated device associated group includes: Based on a preset spatial grid, the real-time temperature data of the target area is discretized to obtain spatial hotspot regions that characterize local thermal disturbance events; A sensitivity analysis method with embedded hotspot area spatial influence attenuation factor is used to construct the sensitivity intensity relationship between air conditioning equipment and spatial hotspot areas; Based on the sensitivity intensity relationship, a clustering method with enhanced local sensitivity is used to group the air conditioning equipment, resulting in equipment association groups.
3. The air conditioning control method based on group control computing according to claim 2, characterized in that, The establishment of the sensitivity intensity relationship between air conditioning equipment and spatial hotspot areas includes: Based on the physical distance between the equipment and the hotspot area and the nonlinear attenuation characteristics of thermal disturbance propagation, a spatial influence attenuation factor is constructed. By combining the spatial influence attenuation factor with the historical response characteristics of the equipment, a sensitivity intensity relationship reflecting the influence of spatial location is generated.
4. The air conditioning control method based on group control computing according to claim 1, characterized in that, The equipment control margin for generating embedded load fluctuation cycle characteristics includes: Extract the periodic peak-valley time difference characteristics of the load changes of each air conditioning unit to determine the load fluctuation period of the unit; The load fluctuation cycle is verified using a load cycle stability verification method to form a verified load fluctuation cycle. Based on the verified load fluctuation cycle and the rated operating regulation capacity of the equipment, a control margin for the equipment embedded with the load fluctuation cycle characteristics is generated.
5. The air conditioning control method based on group control computing according to claim 4, characterized in that, The method for verifying load fluctuation cycles using a load cycle stability verification method includes: Based on historical load fluctuation data from multiple cycles, the differences in the timing of peaks and troughs in each cycle are calculated. A robustness assessment method based on the distribution of peak and valley time differences was adopted to confirm the stability level of each cycle and to sort and screen the cycle stability. Typical load fluctuation cycles are determined from the stability cycles selected through sorting and screening, and used as the verified load fluctuation cycles.
6. The air conditioning control method based on group control computing according to claim 1, characterized in that, The load margin correction coefficient for generating the characterization of equipment health degradation level includes: Extract equipment load response change data and cumulative equipment runtime from the maintenance records of each equipment to construct the equipment load response degradation trajectory. A nonlinear regression method that integrates the preset equipment lifespan expectation and the equipment load response degradation trajectory is used to fit the health degradation rate of the equipment; Equipment health degradation levels are classified according to the rate of health deterioration, and load margin correction coefficients are generated.
7. The air conditioning control method based on group control computing according to claim 6, characterized in that, The nonlinear regression method employing the fusion of pre-defined equipment lifespan expectations and equipment load response degradation trajectories includes: Based on the equipment load response degradation trajectory, a nonlinear degradation fitting function considering the preset equipment life expectation constraint is constructed; A nonlinear fitting method based on least squares support vector machine is used to solve the nonlinear degradation fitting function and obtain the health degradation rate of the equipment.
8. The air conditioning control method based on group control computing according to claim 1, characterized in that, The generated load control command for the air conditioning equipment includes: Based on historical environmental disturbance data and regional load change data, a recursive prediction model with embedded time-series memory mechanism is used to predict environmental heat load disturbances. Based on the predicted environmental heat load disturbance results and equipment association grouping, a negotiation game algorithm is used to determine the initial load compensation requirements of each equipment group; A margin optimization allocation algorithm combining equipment control margin and load margin correction coefficient is adopted to perform collaborative optimization calculation on the initial load compensation demand and generate load control commands for air conditioning equipment.
9. The air conditioning control method based on group control computing according to claim 8, characterized in that, The method of determining the initial load compensation requirements for each equipment group using a negotiation game algorithm includes: A negotiation game framework with environmental heat load disturbance as the external constraint is constructed, and the negotiation payoff function of equipment grouping is defined. Based on the negotiated payoff function among the equipment groups under the game framework, the load compensation demand solution that satisfies the global disturbance demand balance is determined, and the initial load compensation demand of the equipment groups is generated.
10. An air conditioning control system based on group control computing, implemented based on the air conditioning control method based on group control computing according to any one of claims 1-9, characterized in that, The system includes: Sensitive Grouping Module: This module generates equipment association groups based on the real-time operating load data of each air conditioning unit and the real-time monitored temperature data of the target area, using a clustering analysis method that incorporates the sensitivity to local thermal disturbances. Cycle margin module: used to extract the periodic fluctuation pattern of the load based on the historical operating load data of each air conditioning unit, so as to generate the equipment control margin embedded with the load fluctuation cycle characteristics; Deterioration correction module: Used to analyze the maintenance records and cumulative operating time of each air conditioning unit using a regression fitting method that integrates degradation rate perception, and generate load margin correction coefficients that characterize the health degradation level of the equipment; Negotiation and Allocation Module: This module uses a negotiation and allocation algorithm embedded with environmental heat load disturbance prediction to collaboratively calculate equipment association grouping, equipment control margin, and load margin correction coefficient, thereby generating load control commands for air conditioning equipment.