A method and system for adaptive adjustment of operating parameters of a five-constant system

Abstract

The present application relates to the field of data processing and intelligent control, and particularly relates to a five-constant system operation parameter adaptive adjustment method and system, which comprises the following steps: collecting multi-region temperature data of the five-constant system in real time; calculating the inter-region parameter uniformity of each region temperature data; obtaining a system steady-state index by combining the inter-region parameter uniformity with the maximum value of information entropy of each region parameter; constructing an abnormal early warning index based on the steady-state index of different time windows; judging the stability of different length time windows based on the abnormal early warning index, and selecting a decision window; calculating the parameter mean value of each region based on the temperature data of each region in the decision window, obtaining an entropy weight benchmark value by combining the information entropy, calculating the difference between the parameter mean value and the global entropy weight benchmark value as a parameter deviation, and dynamically and adaptively adjusting the operation state of the five-constant system. The present application combines temperature space and time sequence characteristics, dynamically selects a window, and adjusts the partition by entropy weight, thereby improving temperature control uniformity and stability, and taking into account both precision and energy saving.

Description

Technical Field

[0001] This invention relates to the field of data processing and intelligent control. In particular, it relates to a method and system for adaptive adjustment of operating parameters of a five-constant system. Background Technology

[0002] The Five Constant System is a high-end indoor environmental control system that integrates constant temperature, constant humidity, constant oxygen, constant cleanliness, and constant quietness. Through the coordinated work of radiant terminals, fresh air units, temperature and humidity regulation, and air purification modules, it provides a stable and comfortable indoor environment. Its temperature data regulation effect directly determines the environmental comfort, system energy efficiency, and equipment lifespan.

[0003] In actual operation, random disturbances such as outdoor weather fluctuations, changes in indoor personnel and equipment loads, and the opening and closing of doors and windows can cause significant unevenness in heat and humidity loads across different areas, placing high demands on the system's adaptive adjustment capabilities. Scientific temperature data control is key to ensuring environmental balance across multiple areas, reducing energy consumption, and improving system stability.

[0004] Existing five-constant system regulation methods only monitor instantaneous parameters, neglecting the uniformity between regions and the stability over time. This easily leads to problems such as local overheating and overcooling, and frequent temperature fluctuations, resulting in frequent system start-ups and shutdowns, low energy efficiency, slow response, and reduced comfort. At the same time, equal-weighted control is easily affected by local abnormal areas, cannot adapt to dynamic differences in multiple regions, and lacks adaptability and robustness. Summary of the Invention

[0005] To address the problems of existing five-constant system regulation methods that only monitor instantaneous parameters, neglect the uniformity of parameters between regions and temporal stability, and have insufficient control robustness and adaptive capability, this invention provides solutions in the following aspects.

[0006] In a first aspect, a method for adaptive adjustment of operating parameters of a five-constant system includes: acquiring temperature data of multiple regions of the five-constant system in real time; determining the uniformity of parameters between regions based on the temperature data of each region, wherein the uniformity of parameters between regions is used to reflect the degree of dispersion difference of temperature data of multiple regions at the same time; determining the system steady-state index within a time window by combining the uniformity of parameters between regions with the maximum information entropy of parameters in each region; determining an anomaly warning index based on the system steady-state index of different length time windows; judging the system stability according to the anomaly warning index; selecting the corresponding time window as a decision window; determining the mean value of parameters in each region based on the temperature data within the decision window; weighting the mean value of parameters in each region by combining the information entropy of parameters in each region to obtain a global entropy weight benchmark value; using the difference between the mean value of regional parameters and the global entropy weight benchmark value as the parameter deviation; and performing adaptive adjustment on each region of the five-constant system based on the parameter deviation.

[0007] Preferably, the calculation method for the uniformity of parameters between regions is as follows: Calculate the normalized value of the temperature data corresponding to each region at each time. Use the ratio of the discrete characteristics of the normalized temperature data of each region to the average level as the uniformity of the parameters between regions of the five constant systems.

[0008] Preferably, the steady-state index of the system is calculated as follows: The system steady-state index is obtained by calculating the standard deviation of the uniformity of parameters between regions at all times within the time window, calculating the information entropy of the normalized temperature data of all sampling points in each region within the time window, multiplying the maximum value of the information entropy of the temperature data in each region by the standard deviation of the uniformity of parameters between regions, and then performing a negative exponential operation.

[0009] Preferably, the abnormal warning index is calculated as follows: The ratio of the minimum to the mean of the system steady-state index for different time windows is calculated as the stability equilibrium index. The ratio of the standard deviation to the mean of the system steady-state index for different time windows is calculated as the coefficient of variation. The anomaly warning index is obtained by subtracting the stability equilibrium index from 1 and multiplying it by the coefficient of variation.

[0010] Preferably, the mean value of the parameter is calculated as follows: The normalized temperature data of each region within the decision window at each time point is calculated and averaged to obtain the mean parameter value of each region.

[0011] Preferably, the global entropy weight benchmark value is calculated as follows: Calculate the information entropy of temperature data for each region within the decision window, perform a negative exponential operation on the information entropy of the temperature data to obtain the stability weight of each region, perform a weighted operation on the mean parameter of each region and its respective stability weight, and sum the weighted results of all regions to obtain the global weighted temperature value; sum the stability weights of all regions to obtain the global weight value; and use the ratio of the global weighted temperature value to the global weight value as the global entropy weight benchmark value.

[0012] Preferably, the decision window is selected in the following way: A preset abnormal warning index threshold is set. When the abnormal warning index is less than the threshold, the temperature data of the five constant systems is not adjusted. When the abnormal warning index is greater than or equal to the threshold, the time window corresponding to the maximum value of the abnormal warning index is selected as the decision window.

[0013] Secondly, a five-constant system operating parameter adaptive adjustment system includes: a processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, the five-constant system operating parameter adaptive adjustment method described in any one of the claims is implemented.

[0014] The present invention has the following effects: 1. This invention utilizes the spatial dispersion and temporal fluctuation characteristics of temperature in multiple regions, and constructs a steady-state evaluation system based on the dual dimensions of parameter uniformity and temperature information entropy between regions. This effectively compensates for the shortcomings of traditional methods that neglect regional temperature uniformity and long-term operational stability, and significantly improves the uniformity of temperature distribution across the entire region.

[0015] 2. This invention constructs an anomaly warning index by comparing multiple time windows over a long period of time. It can accurately distinguish between instantaneous environmental interference and continuous temperature control anomalies, dynamically select the best decision statistical window, get rid of the limitations of traditional fixed time window control, and adapt to various complex working conditions such as personnel flow, load changes, and seasonal changes. It significantly enhances the system's autonomous judgment and dynamic adaptation capabilities, and makes the control operation more stable.

[0016] 3. This invention constructs a global temperature control benchmark value through an information entropy weighting mode, automatically weakens the data weight of abnormal fluctuation areas, and completes zoned PID adaptive adjustment based on actual temperature deviation. It can accurately match the temperature control needs of different independent control areas, effectively avoid the problems of local over-adjustment and temperature adjustment lag caused by unified control, improve the comfort and control accuracy of constant temperature control, reduce the ineffective action of the actuator, and reduce the energy consumption of the entire five constant systems. Attached Figure Description

[0017] Figure 1 This is a flowchart of steps S1-S5 in an adaptive adjustment method for the operating parameters of a five-constant system according to an embodiment of the present invention.

[0018] Figure 2 This is a structural flowchart of a five-constant system operating parameter adaptive adjustment system according to an embodiment of the present invention. Detailed Implementation

[0019] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are some embodiments of the present invention, but not all embodiments.

[0020] The specific embodiments of the present invention will now be described in detail with reference to the accompanying drawings.

[0021] Reference Figure 1 An adaptive adjustment method for the operating parameters of a five-constant system includes steps S1-S5, as detailed below: S1: Real-time acquisition of temperature data from multiple zones of the five constant systems.

[0022] Temperature sensors are deployed in each independent control area covered by the five constant temperature system to achieve independent temperature monitoring of multiple indoor zones. Temperature parameters of each independent control area are periodically collected in real time at a preset fixed acquisition time interval, exemplified by 30-second intervals. This yields multi-zone temperature monitoring data containing time-series information, providing a data foundation for subsequent analysis of parameter uniformity between zones, stability assessment, and adaptive parameter adjustment.

[0023] After acquiring continuous temperature monitoring data from multiple regions over time, relying solely on instantaneous temperature values ​​from a single moment or region cannot fully reflect the overall control performance of the five constant temperature and humidity system. To accurately identify the spatial uniformity and temporal stability of the temperature field across multiple regions, and to avoid control malfunctions caused by local disturbances, instantaneous fluctuations, or fixed window lag, a comprehensive evaluation of the collected temperature data is necessary to determine a dynamic decision window that truly reflects the system's stable state. The specific steps are as follows: S2: Determine the uniformity of parameters between regions based on the temperature data of each region. The uniformity of parameters between regions is used to reflect the degree of dispersion difference of temperature data in multiple regions at the same time. Combine the uniformity of parameters between regions with the maximum information entropy of parameters in each region to determine the steady-state index of the system within the time window.

[0024] To eliminate the interference of differences in the dimensions and amplitudes of temperature values ​​in different regions on the evaluation results, the temperature data collected from all regions at each acquisition time were first normalized to obtain the normalized temperature values ​​for each region at the corresponding time.

[0025] Furthermore, the standard deviation of the normalized temperature data of all regions at the same time is taken as the data dispersion feature, and the average value of the normalized temperature data of all regions at the same time is taken as the overall average level. The ratio of the standard deviation to the average value is taken as the temperature uniformity index between regions corresponding to the five constant systems, so as to accurately quantify the degree of temperature distribution difference between different independent control regions at the same time.

[0026] Multiple data analysis time windows of different durations are defined, with the time window duration ranging from 5 minutes to 30 minutes. 15 minutes is preferred as the standard analysis duration. The start and end time nodes corresponding to each group of time windows are clearly defined, and all valid temperature sampling data and corresponding calculation results within the time window range are determined.

[0027] For each set of completed time windows, the calculated values ​​of temperature uniformity between regions corresponding to all acquisition times included in the time window are collected. The standard deviation of all collected values ​​of temperature uniformity between regions is calculated. This standard deviation is used to characterize the fluctuation range of the spatial distribution difference of the five constant systems temperature within the time window.

[0028] Extract the normalized temperature data of each independent control area at all sampling times within the current time window, calculate the information entropy of the temperature time series data of each independent control area, use the information entropy to quantify the degree of disorder of temperature fluctuations and changes over time within the corresponding independent control area, compare the information entropy values ​​of all independent control areas one by one, and determine the maximum value among all information entropy values.

[0029] The maximum information entropy value is multiplied by the standard deviation of the temperature uniformity between regions corresponding to the current time window. The result of this product is then subjected to a negative exponentiation to obtain the system steady-state index for the corresponding time window. The system steady-state index can comprehensively evaluate the uniformity of the spatial distribution of the five constant temperature systems within the corresponding time period, as well as the stability of the temperature time series operation within each independent control region.

[0030] Specifically, the system steady-state index satisfies the following relationship: ; In the formula, Indicates time window The corresponding system steady-state index, Indicates time window The standard deviation of the temperature uniformity values ​​between regions at all times within the timeframe. Indicates time window Inner Temperature uniformity index between regions corresponding to each data collection time. This indicates the total number of independent control regions set up in the constant system. Indicates the first Each region within the time window Temperature information entropy of all sampled temperature data. This represents the maximum value in the temperature information entropy across all regions. Represented by natural constant An exponential function with base 0.

[0031] Existing temperature control methods for five constant temperature systems mostly rely solely on instantaneous temperature values ​​for regulation, failing to simultaneously consider two core influencing factors: spatial temperature distribution fluctuations and the intensity of temporal temperature fluctuations in different zones. This can easily lead to problems such as local temperature imbalances, frequent regulation actions, and poor system stability. Therefore, this paper integrates the spatial temperature distribution fluctuation characteristics with the temporal fluctuation characteristics of different zones, and builds a fusion based on mathematical statistical features and information entropy theory to achieve accurate quantitative evaluation of the system's temperature control status within different time windows.

[0032] The standard deviation of temperature uniformity between regions can objectively reflect the fluctuation and change pattern of the spatial distribution of temperature across the entire region, while the regional temperature information entropy can accurately characterize the degree of disorder in the temporal fluctuation of temperature in each zone. These two characteristics correspond to the unstable factors in the spatial and temporal dimensions, respectively, covering all evaluation dimensions of the temperature control stability of the five constant temperature systems. The greater the amplitude of spatial distribution fluctuation and the higher the degree of maximum temporal temperature fluctuation in each zone, the higher the overall temperature control instability of the system. Multiplying the two can simultaneously amplify the negative impact of the dual unstable factors, intuitively reflecting the overall imbalance of the system and conforming to the actual temperature control operation pattern.

[0033] A negative exponential function is used for calculation to keep the system steady-state index within a reasonable range. The higher the degree of system instability, the smaller the calculated steady-state index value; the more uniform the system temperature distribution and the smoother the temporal fluctuations of the zone temperature, the closer the calculated steady-state index value is to the standard high value. The trend of the value change has a positive correlation with the actual steady state of the system, which facilitates subsequent threshold determination and control strategy triggering.

[0034] Since the steady-state index values ​​corresponding to different time windows vary significantly, relying solely on a single steady-state index cannot intuitively determine whether the temperature control conditions of the five constant systems are abnormal, nor can it accurately select the optimal analysis period that is suitable for the current temperature control state.

[0035] To quantify the risk of system temperature control anomalies and determine statistical analysis periods that accurately reflect on-site temperature change patterns, an anomaly warning index is needed to characterize the severity of system temperature control anomalies. This index is used to determine the overall stable operating status of the system. Based on the determination results, a target time period suitable for the current control scenario is selected from multiple time windows of different lengths as the decision window. The specific steps are as follows: S3: Determine the anomaly warning index based on the system steady-state index with different time windows, judge the system stability based on the anomaly warning index, and select the corresponding time window as the decision window.

[0036] After obtaining the system steady-state indices corresponding to multiple time windows of different durations, the minimum value of all system steady-state indices is extracted first, and the arithmetic mean of all system steady-state indices is calculated. The ratio of the minimum steady-state index to the average steady-state index is calculated, and the result is set as the stability equilibrium index. The stability equilibrium index can objectively reflect the equilibrium state of the overall steady-state operation level of the system under various statistical durations, which is convenient for identifying the lower limit level of the system's steady-state performance.

[0037] By integrating all system steady-state indices corresponding to different time windows, and combining all steady-state indices into a complete set of statistical data, the standard deviation of the overall distribution of this dataset is calculated to characterize the overall fluctuation range of the system steady-state evaluation results under various statistical durations.

[0038] The coefficient of variation is obtained by comparing the standard deviation with the average value of the system steady-state index. The coefficient of variation can accurately reflect the data dispersion of the system steady-state index under different statistical durations and intuitively reflect the fluctuation differences of the steady-state evaluation results.

[0039] The anomaly warning index is obtained by subtracting the stability equilibrium index from the numerical value of 1, and then multiplying the difference by the coefficient of variation. This anomaly warning index integrates two major characteristics: the system's steady-state lower limit deviation and overall data fluctuation, comprehensively quantifying the degree of abnormal deviation in the temperature control conditions of the five constant temperature systems. It effectively improves the accuracy of identifying abnormal operating conditions, reduces misjudgments caused by relying on single statistical indicators, and ensures comprehensive and reliable temperature control status assessment results.

[0040] Specifically, the anomaly warning index satisfies the following relationship: ; In the formula, Indicates an abnormal warning index. This represents the minimum value among all system steady-state exponents corresponding to different time windows. This represents the average of the system steady-state exponents corresponding to all different time windows. This represents the coefficient of variation of the steady-state index of the system corresponding to all different time windows.

[0041] Most existing temperature control strategies for five constant systems use fixed time windows for data analysis and parameter tuning. Fixed windows cannot adapt to dynamic operating conditions such as indoor personnel movement, changes in environmental heat exchange, and equipment start-up and shutdown disturbances, which can easily lead to problems such as misjudgment of instantaneous temperature fluctuations and missed judgment of long-term temperature imbalances.

[0042] Meanwhile, a single steady-state index can only reflect the operating status of a single time period, and cannot make a horizontal comparison of the system's stable performance under different statistical durations, making it difficult to distinguish between real temperature control faults and short-term environmental noise interference.

[0043] Therefore, by combining the steady-state data equilibrium degree with the discrete fluctuation characteristics of the data to jointly model, an anomaly early warning index calculation formula is constructed to accurately distinguish between real abnormal working conditions and instantaneous interference signals, and simultaneously complete the automatic selection of the optimal decision window.

[0044] The ratio of the steady-state minimum to the steady-state average is less than 1. The larger the ratio, the more consistent the system's temperature control stability level is under different time windows, and the more homogeneous the overall operating state is. If the overall steady-state value is low at this time, it means that the entire domain is in a long-term low temperature control stability state, and the probability of temperature control anomalies in the system is higher. The smaller the ratio, the more significant the difference between the worst stable period and the normal operating period. If the overall average steady-state level is high, such short-term low data is mostly caused by instantaneous environmental disturbances and does not belong to equipment temperature control failure.

[0045] The coefficient of variation directly reflects the density of the distribution of steady-state indices in each group. The larger the coefficient of variation, the greater the difference in the system's stable state at different times, the more drastic the fluctuations in operating conditions, and the higher the probability of abnormal temperature control.

[0046] The real-time calculated anomaly warning index is compared with a preset threshold of 0.5, which can be adjusted according to specific circumstances. To further explain, the anomaly warning index is obtained by multiplying the steady-state equilibrium deviation by the coefficient of variation. The index usually takes values ​​between 0 and 1, with 0.5 being the critical position in the middle of the range. It can clearly distinguish between the stable operation range and the abnormal deviation range, and the boundary is intuitive and clear, which facilitates program logic writing and rapid status determination.

[0047] When the abnormal warning index is less than the preset threshold, it is determined that the current indoor temperature distribution is uniform, the time-series temperature fluctuation is stable, the system has no temperature control abnormality, and there is no need to perform temperature parameter adjustment. When the abnormal warning index is greater than or equal to the preset threshold, it is determined that the system temperature control status has an abnormal deviation, and the time window corresponding to the minimum value of the system steady-state index is automatically selected as the decision window.

[0048] After determining the system's temperature control operating status and establishing a dedicated decision window, the system has locked in a valid statistical period that aligns with the current actual temperature control conditions. Previously, temperature control methods often directly used a uniform average temperature across the entire region as the adjustment benchmark, failing to distinguish the temperature fluctuation characteristics and operational stability differences of different areas. This easily led to a mismatch between the adjustment benchmark and the actual operating conditions of the region, resulting in localized over-temperature adjustment and uneven control effects.

[0049] To ensure that the temperature control benchmark can accurately adapt to the actual operating status of each independent control area, taking into account both the overall temperature control balance of the entire region and the differentiated fluctuation characteristics of each zone, and to avoid the one-sidedness of control caused by a single average benchmark, the benchmark values ​​are reconstructed by assigning corresponding control weights based on the degree of temperature fluctuation in each region, so that the benchmark values ​​can objectively reflect the overall temperature operation level of the entire region. Therefore, the benchmark values ​​are reconstructed using the effective temperature data within the decision window. The specific steps are as follows: S4: Determine the mean value of parameters for each region based on the temperature data within the decision window, and weight the mean values ​​of parameters by combining the information entropy of parameters for each region to obtain the global entropy weight benchmark value.

[0050] By defining a corresponding effective data statistical period through a predetermined decision window, raw temperature monitoring data collected at all data acquisition times for each independent control area within that statistical period are retrieved. To eliminate calculation deviations caused by different site baseline temperatures and temperature value spans between different areas, and to unify the calculation benchmark for temperature data across the entire region, normalization calculations are performed on the raw temperature data of each independent control area collected at each data acquisition time within the effective data statistical period to determine the normalized temperature value of each independent control area at the corresponding data acquisition time.

[0051] After completing the normalization of all temperature data, data integration calculations are performed separately for each independent control area. The normalized temperature values ​​corresponding to all data acquisition times within the effective data statistical period of each independent control area are summarized. The arithmetic mean of all the normalized temperature values ​​in the same area is calculated to determine the mean regional temperature parameter for each independent control area.

[0052] This data processing method can completely eliminate data interference caused by fluctuations in the ambient temperature, accurately extracting the stable temperature level of each independent control area during the effective data statistical period, and ensuring that the average temperature parameters obtained from each independent control area have a unified quantitative standard. Simultaneously, it provides standardized and consistent basic computational data for subsequent weight allocation and weighted fusion calculations based on information entropy, effectively ensuring that the final global operating benchmark value accurately reflects the actual temperature distribution across the entire region, further improving the accuracy and rationality of subsequent temperature control strategies.

[0053] It should be noted that an independent control zone refers to an indoor functional zoning unit that is pre-divided according to the building's interior space layout, air conditioning ductwork arrangement, and indoor heating and cooling load distribution patterns, and each zone is equipped with an independent temperature acquisition module and an independent temperature control actuator. Each independent control zone can independently complete real-time temperature data acquisition, and can also receive control commands to achieve independent temperature adjustment within the zone. The temperature control operation status of each zone is independent of each other.

[0054] All temperature monitoring data corresponding to each independent control area within the statistical decision window are used to calculate the temperature information entropy value of each independent control area according to the principle of information entropy calculation. The temperature information entropy value is used to objectively characterize the degree of drastic temperature temporal fluctuation within the corresponding independent control area.

[0055] To transform temperature fluctuation characteristics into quantifiable weights that can participate in weighted calculations, a negative exponential operation is performed on the temperature information entropy values ​​obtained from each independent control region. The result is then set as the operational stability weight for the corresponding independent control region. The greater the degree of temperature fluctuation, the higher the corresponding temperature information entropy value, and the smaller the stability weight value obtained after negative exponential conversion. This reflects the lower priority of abnormal fluctuation regions in the global temperature control benchmark. Conversely, the more stable the temperature operation, the higher the obtained stability weight value, and the more significant the reference weight of stable regions in global temperature control.

[0056] The mean temperature parameter for each independent control region is multiplied sequentially by its corresponding operational stability weight, and the sum of all multiplication results constitutes the global temperature-weighted comprehensive statistic. The operational stability weights for all independent control regions are then summed to form the global weighted aggregate metric. The global temperature-weighted comprehensive statistic is divided by the global weighted aggregate metric to obtain the global entropy weight baseline value.

[0057] By combining the actual stable temperature operation status of different independent control zones, adaptive weight allocation is achieved, abandoning the traditional average calculation mode of equal value taking. This effectively weakens the interference of abnormal temperature fluctuation areas on the overall control benchmark and strengthens the reference and leading role of temperature-stable areas. The resulting global entropy weight benchmark value can truly reflect the actual operation of the comprehensive indoor temperature control, providing a scientific and reasonable core reference standard for subsequent zone temperature deviation judgment and fine-tuned adaptive adjustment, effectively improving the overall temperature balance control effect and reducing the frequency of ineffective system adjustments.

[0058] Specifically, the global entropy weight benchmark value satisfies the following relationship: ; In the formula, This represents the global entropy weight baseline value. This indicates the decision window that has been selected and confirmed after the abnormal operating conditions have been determined. This represents the total number of independent control regions obtained by dividing the system as a whole. Indicates the first Each independent control area in the decision window The information entropy of all temperature data within the region is used to characterize the severity of temperature fluctuations over time within the independent control area. A higher temperature information entropy value corresponds to more pronounced temperature fluctuations in the independent control area, lower reference value of the average temperature parameters in that area, a smaller temperature stability weight obtained through negative exponential conversion, and a lower proportion of the data from that area participating in the calculation of the global benchmark value. Conversely, a lower temperature information entropy value corresponds to a more stable temperature operation in the independent control area, stronger representativeness of the average temperature parameters in that area, a larger calculated temperature stability weight, and a higher reference proportion of the data from that area in the calculation of the global benchmark value. Indicates the first Each independent control area in the decision window The mean of the internal temperature data, Represented by natural constant An exponential function with base 0.

[0059] In actual operation scenarios, the temperature in some independently controlled areas fluctuates frequently due to factors such as indoor occupant activity, ventilation disturbances, uneven heating and cooling loads, and differences in air supply. The average temperature of such areas with large fluctuations is not representative and cannot truly reflect the normal indoor temperature requirements. If it is included in the calculation as data from stable operating areas, it will cause a deviation in the overall control benchmark value, leading to problems such as overall temperature control imbalance, excessive local adjustment, and frequent system adjustments. Therefore, the temperature fluctuation characteristics of the zones are quantified by information entropy theory, the fluctuation intensity is converted into a quantitative weight, an entropy-weighted average calculation formula is constructed, and reference weights are adaptively allocated based on the temperature operation stability of the zones to complete the accurate solution of the overall temperature control benchmark value.

[0060] The global entropy weight benchmark value integrates the overall temperature level of the entire region with the temperature fluctuation characteristics of each independent control area, and can serve as a unified reference standard for temperature regulation across the entire region. In contrast, previous regulation methods simply issued adjustment commands based on a fixed set temperature without considering the deviation of the actual operating average of each independent control area from the comprehensive benchmark value to formulate differentiated adjustment schemes. This easily leads to problems such as uniform adjustment intensity, inability to adapt to the actual temperature control needs of each zone, and difficulty in achieving balanced and stable temperature control across the entire region.

[0061] To accurately determine the deviation between the current temperature operating status of each independent control zone and the optimal control benchmark, and to match the corresponding adjustment intensity based on the deviation magnitude to achieve precise temperature control in each zone, deviation calculation is performed based on the determined average temperature parameters of the zones and the global entropy weight benchmark value. The obtained deviation values ​​are then combined to complete the zoned adaptive temperature adjustment of the five constant temperature system. The specific implementation method is as follows: S5: The difference between the mean value of the regional parameters and the global entropy weight benchmark value is used as the parameter deviation, and adaptive adjustment is performed on each region of the five constant systems based on the parameter deviation.

[0062] After obtaining the temperature parameter deviations corresponding to each independent control zone, the PID control algorithm is used to solve for the real-time adjustment change of the end temperature control actuator in the kth independent control zone. The specific calculation formula is as follows: ; In the formula, Indicates the first Each independent control area Changes in motion adjustment over time This represents the proportional control coefficient. Indicates the first Real-time temperature parameter deviation in each independent control area Represents the integral control coefficient. These represent the differential control coefficients. All three types of coefficients are preset system control hyperparameters, and their values ​​can be flexibly selected based on the on-site operating environment, space heat exchange conditions, and actual commissioning experience. In this embodiment, the proportional control coefficient is preferably set to 0.5, the integral control coefficient to 0.1, and the differential control coefficient to 0.2. Alternatively, a fuzzy PID parameter adaptive tuning method can be used to achieve dynamic optimization and matching of the three types of control coefficients, further improving parameter adaptability.

[0063] It should be noted that the adjustment change can be directly converted into the actual operating adjustment of various temperature control actuators. For example, when the controlled component is an electric two-way regulating valve, the adjustment change corresponds to the valve opening adjustment percentage; when the controlled component is a regulating fan, the adjustment change corresponds to the fan operating speed adjustment percentage.

[0064] By using a PID algorithm to synchronously complete proportional rapid correction, integral steady-state error elimination, and derivative advance prediction adjustment based on real-time temperature deviation, dynamic closed-loop correction of temperature deviation in each independent control area can be achieved. This not only enables rapid response to instantaneous temperature fluctuations but also gradually eliminates long-term static temperature control errors, effectively suppressing overshoot and oscillation in system temperature control. It ensures that the temperature in each independent control area steadily approaches the global entropy weight benchmark value, significantly improving the response speed and constant temperature control accuracy of zoned temperature control, while reducing equipment wear and energy consumption caused by frequent and large-amplitude movements of the end actuator.

[0065] This invention also provides an adaptive adjustment system for the operating parameters of a five-constant system. For example... Figure 2 As shown, the system includes a processor and a memory. The memory stores computer program instructions, which, when executed by the processor, implement a machine learning-based adaptive adjustment method for the operating parameters of a five-constant system according to the first aspect of the present invention. The system also includes other components well-known to those skilled in the art, such as a communication bus and a communication interface; their configurations and functions are known in the art and will not be described further here.

[0066] It should be noted that those skilled in the art can make various modifications and improvements without departing from the inventive concept, and these all fall within the scope of protection of this invention. Therefore, the scope of protection of this patent should be determined by the appended claims.

Claims

1. A method for adaptive adjustment of operating parameters of a five-constant system, characterized in that, include: Real-time acquisition of temperature data from multiple zones of the five constant temperature system; Based on the temperature data of each region, the uniformity of parameters between regions is determined. The uniformity of parameters between regions is used to reflect the degree of discrete difference of temperature data in multiple regions at the same time. Combining the uniformity of parameters between regions with the maximum information entropy of parameters in each region, the steady-state index of the system within the time window is determined. This includes: calculating the standard deviation of the uniformity of parameters between regions at all times within the time window; calculating the information entropy of normalized temperature data of all sampling points in each region within the time window; and multiplying the maximum information entropy of temperature data in each region by the standard deviation of the uniformity of parameters between regions and then performing a negative exponential operation to obtain the steady-state index of the system. The anomaly warning index is determined based on the system steady-state index of different time windows, including: calculating the ratio of the minimum value to the mean value of the system steady-state index of different time windows as the stability equilibrium index; calculating the ratio of the standard deviation to the mean value of the system steady-state index of different time windows as the coefficient of variation; and multiplying the stability equilibrium index by the coefficient of variation after subtracting the coefficient of variation from 1 to obtain the anomaly warning index. The system stability is determined based on the anomaly warning index, and a corresponding time window is selected as the decision window. The mean values ​​of parameters in each region are determined based on the temperature data within the decision window. The mean values ​​of these parameters are then weighted according to the information entropy of the parameters in each region to obtain a global entropy weight benchmark value. This process includes: calculating the information entropy of the temperature data in each region within the decision window; performing a negative exponential operation on the information entropy of the temperature data to obtain the stability weight of each region; weighting the mean values ​​of the parameters in each region with their respective stability weights; summing the weighted results of all regions to obtain the global weighted total temperature value; summing the stability weights of all regions to obtain the global weight total value; and using the ratio of the global weighted total temperature value to the global weight total value as the global entropy weight benchmark value. The difference between the mean value of the regional parameters and the global entropy weight benchmark value is used as the parameter deviation, and adaptive adjustment is performed on each region of the five constant systems based on the parameter deviation.

2. The adaptive adjustment method for the operating parameters of a five-constant system according to claim 1, characterized in that, The calculation method for the uniformity of parameters between regions is as follows: Calculate the normalized value of the temperature data corresponding to each region at each time. Use the ratio of the discrete characteristics of the normalized temperature data of each region to the average level as the uniformity of the parameters between regions of the five constant systems.

3. The adaptive adjustment method for the operating parameters of a five-constant system according to claim 1, characterized in that, The mean value of the parameter is calculated as follows: The normalized temperature data of each region within the decision window at each time point is calculated and averaged to obtain the mean parameter value of each region.

4. The adaptive adjustment method for the operating parameters of a five-constant system according to claim 1, characterized in that, The decision window is selected in the following way: A preset abnormal warning index threshold is set. When the abnormal warning index is less than the threshold, the temperature data of the five constant systems is not adjusted. When the abnormal warning index is greater than or equal to the threshold, the time window corresponding to the maximum value of the abnormal warning index is selected as the decision window.

5. A five-constant system operating parameter adaptive adjustment system, characterized in that, include: A processor and a memory, wherein the memory stores computer program instructions that, when executed by the processor, implement the adaptive adjustment method for the operating parameters of a five-constant system according to any one of claims 1-4.

Images