Heat exchange system optimization method and system based on big data analysis

Abstract

The present invention discloses a heat exchange system optimization method and system based on big data analysis, which belongs to the field of heat exchange system optimization processing technology, and includes the following steps: analyzing the computing resource demand evaluation index of each data stream, thereby performing resource restrictions, analyzing the real-time data flow fluctuation coefficient and the system computing power evaluation index based on the computing resource demand evaluation index of each data stream, adjusting the sampling frequency of each data stream, and predicting the heat load of the heat exchange system, thereby performing dynamic adjustment of the heat exchange system. By evaluating the computing resource demand and performing reasonable resource restrictions, the present invention avoids delays or errors caused by resource competition during data processing of different data streams, ensures the stability and efficiency of each data stream processing, improves the quality of data acquisition and resource utilization efficiency, optimizes resource allocation, improves energy utilization efficiency, reduces energy consumption, and improves the stability, accuracy and economy of the heat exchange system operation.

Description

Technical Field

[0001] The present invention relates to the technical field of heat exchange system optimization processing, and in particular to a heat exchange system optimization method and system based on big data analysis. Background Art

[0002] With the intensifying energy crisis and increasingly serious environmental pollution, improving energy efficiency and reducing energy consumption have become key issues. In the industrial sector, heat exchange systems are crucial for energy recovery and utilization. Heat load forecasting, a key technology for optimizing heat exchange system operations, predicts the heat load required by the system over a period of time, helping the system adjust operating parameters in advance, optimize resource allocation, and improve system efficiency and cost-effectiveness.

[0003] Currently, the heating supply is predicted based on the target temperature, and the random forest algorithm is used to screen and reduce the dimensionality of the features; then the data is standardized; and then a heat load prediction model based on the long-term and short-term time series network is established. The model captures long-term and short-term feature information through convolutional layers and cyclic layers, and then introduces the concept of cyclic skip layers to capture longer-term feature information. At the same time, the autoregressive algorithm is used to add linear processing capabilities to the model, enhancing the robustness of the model.

[0004] However, existing heat exchange systems, such as those used for heat load prediction, face challenges with data collection timeliness and accuracy due to fluctuations in data volume. Furthermore, data processing suffers from an imbalance in computing resource consumption across different data streams, which can easily lead to interference between multiple tasks. This results in inefficient processing of large amounts of data and can even lead to data processing errors, compromising the accuracy of heat load predictions and optimizing the overall heat exchange system. Summary of the Invention

[0005] In order to overcome the problems existing in the prior art, the first aspect of the present invention provides a heat exchange system optimization method based on big data analysis, including the following steps: S1, recording each heat load prediction data stream of the heat exchange system as each data stream, performing computing resource demand evaluation of each data stream, analyzing the computing resource demand evaluation index of each data stream, and thereby performing resource restriction.

[0006] S2, based on the computing resource demand evaluation indicators of each data flow, and synchronously analyzing the real-time data flow fluctuation coefficient and system computing power evaluation indicators, generates adaptive sampling frequency control instructions to adjust the sampling frequency of each data flow.

[0007] S3, collecting data based on the adjusted sampling frequency, thereby predicting the heat load of the heat exchange system.

[0008] S4, dynamically adjusting the heat exchange system according to the heat load prediction result of the heat exchange system.

[0009] A second aspect of the present invention provides a heat exchange system optimization system based on big data analysis, comprising:

[0010] The computing resource demand assessment module is used to record the heat load prediction data streams of the heat exchange system as data streams, perform computing resource demand assessment on each data stream, analyze the computing resource demand assessment indicators of each data stream, and thus perform resource restrictions.

[0011] The sampling frequency adjustment module is used to generate adaptive sampling frequency control instructions based on the computing resource demand evaluation indicators of each data stream, and synchronously analyze the real-time data flow fluctuation coefficient and system computing power evaluation indicators to adjust the sampling frequency of each data stream.

[0012] The heat load prediction module is used to collect data based on the adjusted sampling frequency, thereby predicting the heat load of the heat exchange system.

[0013] The heat exchange system dynamic adjustment module is used to dynamically adjust the heat exchange system according to the heat load prediction results of the heat exchange system.

[0014] One or more technical solutions provided in the present invention have at least the following technical effects or advantages:

[0015] 1. The heat exchange system optimization method based on big data analysis provided by the present invention, by evaluating the computing resource requirements and making reasonable resource restrictions, avoids delays or errors caused by resource competition during data processing of different data streams, and ensures the stability and efficiency of each data stream processing. At the same time, the sampling frequency is adjusted according to the computing resource demand evaluation index, the real-time data flow fluctuation coefficient and the system computing power evaluation index. It can not only increase the sampling frequency when the data changes rapidly to obtain more accurate and timely data, but also reduce the sampling frequency when the data is stable to avoid resource waste, effectively improving the quality of data acquisition and resource utilization efficiency. Based on the adjusted sampling frequency, data acquisition and heat load prediction are carried out, and the system is dynamically adjusted according to the prediction results to optimize resource allocation, improve energy utilization efficiency, reduce energy consumption, and improve the stability, accuracy and economy of the heat exchange system operation.

[0016] 2. This invention analyzes the computational resource requirements of each data stream and applies resource constraints accordingly. Precise resource constraints can, on the one hand, avoid disordered competition and irrational consumption of system resources, prevent processing delays or errors due to insufficient resources, and improve data processing efficiency and accuracy. Furthermore, it helps improve the operational stability of the entire system. Different data streams operate in their own dedicated resource environments, reducing mutual interference. This makes the system more reliable when processing complex tasks and large amounts of data, providing a solid data foundation for subsequent key processes such as heat load prediction and system dynamic adjustment, ultimately optimizing the overall performance of the heat exchange system.

[0017] 3. The present invention adjusts the sampling frequency of each data stream. When the data of the data stream changes rapidly, increasing the sampling frequency can obtain richer and more timely data, ensure the timeliness and accuracy of the data, provide more accurate data support for heat load prediction, and thus improve the accuracy of the prediction. When the data changes of the data stream are relatively stable, reducing the sampling frequency can avoid unnecessary waste of resources, reasonably optimize the system resource configuration, reduce the system operation burden, and improve the overall operation efficiency of the system.

[0018] 4. The present invention dynamically adjusts the heat exchange system according to the heat load prediction results of the heat exchange system, thereby effectively improving energy utilization efficiency, reducing system operating costs, and helping to maintain the stability of system operation. It can adjust the system status in time according to environmental changes and fluctuations in heat load demand, reduce system operation fluctuations caused by sudden load changes, extend equipment service life, and reduce equipment maintenance costs. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 Flowchart of the heat exchange system optimization method based on big data analysis provided by the present invention.

[0020] Figure 2 This is a schematic diagram of the structure of the heat exchange system optimization system based on big data analysis provided by the present invention. DETAILED DESCRIPTION

[0021] The technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0022] Reference Figure 1 As shown, the first aspect of the present invention provides a heat exchange system optimization method based on big data analysis, comprising the following steps:

[0023] S1, record each heat load prediction data flow of the heat exchange system as each data flow, perform computing resource demand evaluation of each data flow, analyze the computing resource demand evaluation index of each data flow, and thus perform resource restriction.

[0024] In this embodiment, each heat load prediction data stream of the heat exchange system specifically includes: heat load historical data of the heat exchange system, heat exchange system operation data, and meteorological data of the area to which the heat exchange system belongs.

[0025] In this embodiment, the computing resource requirement evaluation index of each data flow is analyzed to implement resource restriction. The specific process is as follows:

[0026] Obtain historical resource parameters for each data flow and analyze the computing resource demand assessment indicators for each data flow.

[0027] The demand quantities corresponding to the demand evaluation index intervals stored in the database are extracted, and the demand quantities corresponding to the intervals in which the computing resource demand evaluation indexes of each data flow are located are mapped and extracted, and recorded as the computing resource demand quantities of each data flow.

[0028] Resource restrictions are imposed based on the computing resource requirements of each data flow.

[0029] It should be noted that the computing resource requirements refer to the number of allocated CPU cores and memory.

[0030] In one specific embodiment, taking the data stream of historical heat load data for a heat exchange system as an example, assume that during multiple past heat load forecasts, the computing resource requirement assessment index for this data stream was 8. The corresponding computing resource requirements extracted from the database indicate the need for an additional 8 CPU cores and 16GB of memory. Based on this, the system imposes resource restrictions on this data stream, allocating these 8 CPU cores and 16GB of memory specifically to the task of processing the data stream of historical heat load data for the heat exchange system.

[0031] In a specific embodiment, by implementing resource restrictions, other data streams cannot occupy these allocated computing resources, thereby ensuring that each data stream has sufficient and stable computing resource support in the subsequent data processing process, avoiding data processing delays or errors caused by resource competition, and ensuring that related operations such as heat load prediction can be carried out efficiently and accurately.

[0032] It should be noted that if the computing resources occupied by a data flow during actual calculation exceed the set computing resource requirements, the calculation task will be stopped.

[0033] In this embodiment, the computing resource demand evaluation index of each data flow is analyzed. The specific analysis process is as follows:

[0034] Obtain historical resource parameters for each data stream, including the average historical data volume, average historical data processing frequency, average historical data type number, and average historical data transmission bandwidth of each data stream.

[0035] It should be noted that the historical resource parameters of each data stream refer to the average value of the resource parameters recorded in each previous heat load prediction of each data stream.

[0036] It should be understood that the historical resource parameters of each data stream are collected from the system program log, where the number of data types refers to the total number of data types contained in the data stream. In a specific embodiment, the data types include but are not limited to numeric, character, and date types.

[0037] Get the reference history resource parameters.

[0038] The reference historical resource parameters include a reference historical data volume average, a reference historical data processing frequency average, a reference historical data type number average, and a reference historical data transmission bandwidth average.

[0039] Based on the historical resource parameters of each data flow and the reference historical resource parameters, a comprehensive analysis and processing is performed to obtain the computing resource demand evaluation indicators of each data flow.

[0040] The computing resource demand evaluation index of each data flow represents the quantitative result of the impact of the historical data volume average, historical data processing frequency average, historical data type number average and historical data transmission bandwidth average of each data flow on the degree of computing resource demand of the data flow.

[0041] In a specific embodiment, the computing resource requirement evaluation index of each data flow is obtained in the following manner:

[0042]

[0043] Among them, A i is the computing resource requirement evaluation index of the i-th data flow, a i is the historical data volume average of the i-th data stream, b i is the mean historical data processing frequency of the i-th data stream, c i is the mean number of historical data types of the i-th data stream, d i is the mean historical data transmission bandwidth of the i-th data stream, a0 is the reference historical data volume mean, b0 is the reference historical data processing frequency mean, c0 is the reference historical data type number mean, d0 is the reference historical data transmission bandwidth mean, x1 is the weight of the historical data volume mean, x2 is the weight of the historical data processing frequency mean, x3 is the weight of the historical data type number mean, x4 is the weight of the historical data transmission bandwidth mean, i is the number of each data stream, i = 1, 2, ..., n, n is the number of data streams.

[0044] It should be noted that the value range of the mean weight of historical data volume, the mean weight of historical data processing frequency, the mean weight of historical data type number and the mean weight of historical data transmission bandwidth are all between 0 and 1. When used, they can be directly extracted from the database to obtain the pre-set value. The specific extraction method is, for example: constructing a mapping set with the mean value of historical data volume, the mean value of historical data processing frequency, the mean value of historical data type number and the mean value of historical data transmission bandwidth respectively and the corresponding weights. When used, the obtained mean value of historical data volume, the mean value of historical data processing frequency, the mean value of historical data type number and the mean value of historical data transmission bandwidth are input into the mapping set, so as to extract the mean weight of historical data volume, the mean weight of historical data processing frequency, the mean weight of historical data type number and the mean weight of historical data transmission bandwidth.

[0045] It should also be noted that the computational resource requirement assessment indicators for each data stream, derived from the analysis and processing of historical resource parameters for each data stream, take into account the interdependence between these parameters. For example, when the mean historical data volume is large, the mean historical data processing frequency often needs to be increased to ensure processing timeliness. A larger mean number of historical data types increases data diversity and complexity, which, for the same number of records, will increase the mean historical data volume. A high mean historical data processing frequency requires a sufficiently large mean historical data transmission bandwidth, otherwise processing efficiency will be affected. Furthermore, a large mean number of historical data types leads to more complex data transmission protocols and encoding methods, which also places higher requirements on the mean historical data transmission bandwidth.

[0046] S2, based on the computing resource demand evaluation indicators of each data flow, and synchronously analyzing the real-time data flow fluctuation coefficient and system computing power evaluation indicators, generates adaptive sampling frequency control instructions to adjust the sampling frequency of each data flow.

[0047] In this embodiment, the real-time data flow fluctuation coefficient is analyzed in detail as follows:

[0048] Extract the initial sampling frequency and data collection verification duration of each data stream.

[0049] During the data collection and verification period, data collection of each data stream is performed at the initial sampling frequency of each data stream, thereby obtaining the data fluctuation parameters of each data stream.

[0050] The data fluctuation parameters of each data flow include the average arrival interval of data packets of each data flow, the total number of data packets, the peak data rate and the average data rate.

[0051] It should be noted that the data fluctuation parameters for each data stream can be obtained through system program log collection. The average packet arrival interval refers to the average time interval between consecutive arriving packets. Shorter intervals indicate faster data flow and potentially more severe fluctuations. The peak data rate refers to the highest rate in the data stream during the data collection and verification period. Higher peak rates may indicate greater fluctuations.

[0052] Extract the reference data fluctuation parameters stored in the database, including the reference data packet average arrival interval, the reference total number of data packets, the reference peak data rate, and the reference average data rate.

[0053] The real-time data flow fluctuation coefficient of each data flow is obtained by analyzing and processing the data fluctuation parameters of each data flow.

[0054] The real-time data flow fluctuation coefficient of each data flow represents a quantitative result of the degree of influence of the average arrival interval of data packets, the total number of data packets, the peak data rate and the average data rate of each data flow on the data flow fluctuation state.

[0055] In a specific embodiment, the real-time data flow fluctuation coefficient of each data flow is obtained in the following manner:

[0056]

[0057] Among them, B i is the real-time data flow fluctuation coefficient of the i-th data stream, α i is the average arrival interval of data packets of the i-th data flow, β i is the total number of data packets in the i-th data flow, γ i is the peak data rate of the i-th data stream, δ i is the average data rate of the i-th data flow, α0 is the reference average data packet arrival interval, β0 is the reference total number of data packets, γ0 is the reference peak data rate, δ0 is the reference average data rate, y1 is the weight of the average data packet arrival interval, y2 is the weight of the total number of data packets, y3 is the weight of the peak data rate, y4 is the weight of the average data rate, i is the number of each data flow, i = 1, 2, ..., n, n is the number of data flows, and e is a natural constant.

[0058] It should be noted that the weight of the average arrival interval of data packets, the weight of the total number of data packets, the weight of the peak data rate and the weight of the average data rate all have value ranges between 0 and 1. When used, the pre-set values ​​can be directly extracted from the database. The specific extraction method can be to construct a mapping set one by one with the average arrival interval of data packets, the total number of data packets, the peak data rate and the average data rate and the corresponding average arrival interval weight of data packets, the total number of data packets, the peak data rate weight and the average data rate weight respectively. When used, the average arrival interval of data packets, the total number of data packets, the peak data rate and the average data rate obtained in real time are input into the corresponding mapping set, so as to extract the weight of the average arrival interval of data packets, the total number of data packets, the peak data rate weight and the average data rate weight.

[0059] It should be noted that the real-time data flow fluctuation coefficients for each data flow are derived by analyzing and processing the data fluctuation parameters of each data flow. This takes into account the correlation between these parameters. For example, a shorter average packet arrival interval indicates an increase in the amount of data arriving per unit time. In this case, the total number of packets typically increases. Simultaneously, the data transmission rate fluctuates more frequently, potentially increasing the peak data rate and, consequently, the average data rate. Conversely, a longer average packet arrival interval slows or even decreases the growth rate of the total number of packets, resulting in relatively stable data transmission and a decrease in both the peak and average data rates. Furthermore, an increase in the total number of packets, while maintaining constant transmission bandwidth and other conditions, can lead to data transmission congestion, lengthening the average packet arrival interval, which in turn affects the peak and average data rates, resulting in peak rate constraints and average rate fluctuations. A higher peak data rate often indicates extremely rapid data transmission during a specific period, potentially significantly shortening the average packet arrival interval during that period. This, in turn, drives a rapid increase in the total number of packets within a short period, leading to a corresponding increase in the average data rate.

[0060] In this embodiment, the system computing power evaluation index and the specific analysis method are as follows:

[0061] The system's CPU idle rate, remaining memory, and GPU idle rate are extracted from the system program log and analyzed to obtain the system computing power evaluation indicators.

[0062] It should be noted that the CPU idle rate refers to the proportion of CPU computing power that is currently not in use, the remaining memory capacity refers to the unoccupied memory capacity in the system, and the GPU idle rate refers to the proportion of GPU computing power that is currently not in use.

[0063] The system computing power evaluation index represents the quantitative result of the degree to which the system's CPU idle rate, memory remaining amount, and GPU idle rate jointly affect the system's remaining computing power status.

[0064] In a specific embodiment, the system computing power evaluation index is obtained in the following manner:

[0065] Extract the reference CPU idle rate, reference memory remaining amount, and reference GPU idle rate stored in the database.

[0066] The system's CPU idle rate, remaining memory, and GPU idle rate are normalized. The system's CPU idle rate is divided by the reference CPU idle rate to obtain a normalized CPU idle rate value. Similarly, the system's remaining memory is divided by the reference remaining memory to obtain a normalized memory remaining value. Finally, the system's GPU idle rate is divided by the reference GPU idle rate to obtain a normalized GPU idle rate value. Next, pre-set CPU idle rate weights, memory remaining weights, and GPU idle rate weights are extracted from the database. The normalized values ​​of the CPU idle rate, memory remaining weight, and GPU idle rate are then multiplied by their corresponding weights: the normalized CPU idle rate is multiplied by the CPU idle rate weight, the normalized memory remaining weight is multiplied by the memory remaining weight, and the normalized GPU idle rate is multiplied by the GPU idle rate weight. These multiplication results are then summed to obtain a composite value. Finally, this composite value is exponentially raised to the base e, and the result of this exponential operation is used as the system computing power evaluation indicator.

[0067] In this embodiment, the system computing power evaluation index is specifically expressed as follows:

[0068]

[0069] Among them, C is the system computing power evaluation index, ε is the system's CPU idle rate, ζ is the system's memory remaining amount, η is the system's GPU idle rate, ε0 is the reference CPU idle rate, ζ0 is the reference memory remaining amount, η0 is the reference GPU idle rate, z1 is the CPU idle rate weight, z2 is the memory remaining amount weight, z3 is the GPU idle rate weight, and e is a natural constant.

[0070] It should be understood that the CPU idle rate weight, memory remaining amount weight and GPU idle rate weight can be directly extracted from the database. For example, the extraction method is to construct a mapping set with the CPU idle rate, memory remaining amount and GPU idle rate and the CPU idle rate weight, memory remaining amount weight and GPU idle rate weight respectively. When used, the CPU idle rate, memory remaining amount and GPU idle rate obtained in real time are input into the mapping set to obtain the CPU idle rate weight, memory remaining amount weight and GPU idle rate weight.

[0071] In this embodiment, the sampling frequency of each data stream is adjusted. The specific analysis process is as follows:

[0072] The first correction factor of the sampling frequency of each data stream is extracted according to the computing resource demand assessment index of each data stream. The specific extraction method is: extract the first correction factor of each sampling frequency corresponding to the interval of each computing resource demand assessment index stored in the database, and map the extracted first correction factor of the sampling frequency corresponding to the interval in which the computing resource demand assessment index of each data stream is located, and record it as the first correction factor of the sampling frequency of each data stream.

[0073] It should be noted that the larger the computing resource demand assessment index is, the higher the computing resource demand assessment index is, which means that the demand for computing resources during data stream processing is higher. In order to reduce the computing power consumption of the system, the sampling frequency should be smaller, and the corresponding extracted first correction factor of the sampling frequency should be smaller.

[0074] The second correction factor of the sampling frequency is extracted according to the system computing power evaluation index. The specific extraction method is: extract the second correction factor of each frequency corresponding to each system computing power evaluation index interval stored in the database, and map the second correction factor of the frequency corresponding to the interval in which the system computing power evaluation index is located, and record it as the second correction factor of the sampling frequency.

[0075] It's important to note that the higher the system computing power evaluation index, the more available computing power the system currently has for data processing. To maximize system resource utilization while ensuring the timeliness and accuracy of data collection, the corresponding second correction factor for the sampling frequency should be larger. This is because a higher sampling frequency captures more data. Assuming the system computing power is sufficient, processing this additional data will not overburden the system, but will instead help improve data processing accuracy and reliability.

[0076] According to the real-time data flow fluctuation coefficient of each data stream, the first correction factor of the sampling frequency of each data stream and the second correction factor of the sampling frequency of each data stream, the sampling adjustment index of each data stream is obtained through comprehensive analysis and processing.

[0077] The sampling adjustment index of each data stream represents the quantitative result of the influence of the real-time data flow fluctuation coefficient of each data stream, the first sampling frequency correction factor of each data stream and the second sampling frequency correction factor of each data stream on the degree of sampling frequency adjustment demand of each data stream.

[0078] The sampling adjustment index of each data stream is obtained as follows: the real-time data flow fluctuation coefficient of each data stream, the first correction factor of the sampling frequency of each data stream and the second correction factor of the sampling frequency are multiplied, and the final product result is used as the sampling adjustment index of each data stream.

[0079] In a specific embodiment, the sampling adjustment index of each data stream is specifically expressed as follows:

[0080]

[0081] Among them, D i is the sampling adjustment index of the i-th data stream, B i is the real-time data flow fluctuation coefficient of the i-th data stream, is the first correction factor of the sampling frequency of the i-th data stream, σ C is the second correction factor of the sampling frequency, i is the number of each data stream, i=1,2,...,n, and n is the number of data streams.

[0082] Extract the preset sampling adjustment indicator threshold in the database.

[0083] The sampling adjustment index of each data stream is subtracted from the sampling adjustment index threshold to obtain the sampling adjustment deviation index of each data stream.

[0084] It should be understood that the sampling adjustment deviation indicator of each data stream may be greater than zero, less than zero, or equal to zero.

[0085] The sampling frequency adjustment value of each data stream is extracted according to the sampling adjustment deviation index of each data stream.

[0086] In a specific embodiment, the extraction method is as follows: extract the sampling frequency adjustment value corresponding to each sampling adjustment deviation index interval stored in the database, and map and extract the sampling frequency adjustment value corresponding to the interval in which the sampling adjustment deviation index of each data stream is located, and record it as the sampling frequency adjustment value of each data stream.

[0087] It should be understood that when the sampling adjustment deviation index of the data stream is greater than zero, it means that after comprehensive analysis, the data of the data stream changes rapidly, and the existing sampling frequency cannot meet the data processing requirements. It is necessary to increase the sampling frequency to obtain more accurate and timely data. At this time, the corresponding extracted sampling frequency adjustment value should be greater than zero, and the larger the sampling adjustment deviation index, the larger the extracted sampling frequency adjustment value should be, so as to more effectively adapt to the changes in the data stream and ensure the quality of data acquisition and the accuracy of subsequent heat load prediction.

[0088] When the sampling adjustment deviation index of the data stream is less than zero, it means that after comprehensive analysis, the data changes of the data stream are relatively stable, and the existing sampling frequency of the data stream exceeds the actual requirements of system resources and data processing, which may cause unnecessary waste of resources. At this time, the corresponding extracted sampling frequency adjustment value should be less than zero, and the larger the absolute value of the sampling adjustment deviation index, the larger the absolute value of the extracted sampling frequency adjustment value should be, so as to reduce the sampling frequency, reasonably optimize the system resource configuration, and reduce resource waste.

[0089] The sampling frequency of each data stream is adjusted according to the initial sampling frequency of each data stream and the sampling frequency adjustment value of each data stream.

[0090] In one specific embodiment, taking a data stream of historical heat load data from a heat exchange system as an example, assume that the initial sampling frequency of the data stream is 10 Hz. The preset sampling adjustment index threshold in the database is 5. After analysis, the sampling adjustment index of this data stream is 8. Subtracting the sampling adjustment index threshold from the sampling adjustment index (i.e., 8-5=3) yields a sampling adjustment deviation index of 3 for this data stream.

[0091] Assuming that the sampling frequency adjustment value corresponding to the sampling adjustment deviation index 3 stored in the database is an increase of 2 Hz, then according to the initial sampling frequency of the data stream and the sampling frequency adjustment value, the sampling frequency is adjusted to 12 Hz (ie 10+2=12).

[0092] S3, collecting data based on the adjusted sampling frequency, thereby predicting the heat load of the heat exchange system.

[0093] In this embodiment, the heat load prediction of the heat exchange system is performed, and the specific analysis method is as follows:

[0094] During the preset monitoring period, data is collected based on the adjusted sampling frequency to obtain the historical heat load data and operation data of the heat exchange system.

[0095] The heat exchange system heat load historical data includes the historical heat load average and the historical heat load change rate of the heat exchange system.

[0096] It should be understood that the historical heat load average and the historical heat load change rate influence each other and are closely related. When the historical heat load change rate is positive and large, it indicates that the heat load is rising rapidly, which will cause the subsequent historical heat load average to gradually increase. Conversely, if the historical heat load change rate is negative and the absolute value is large, it means that the heat load is decreasing rapidly, and the historical heat load average will decrease accordingly.

[0097] The operating data of the heat exchange system include the average temperature difference between the inlet and outlet of the heat exchanger, the average pressure difference between the inlet and outlet of the heat exchanger, and the heat exchange efficiency of the heat exchanger.

[0098] It should be understood that the heat exchange efficiency of the heat exchanger refers to the ratio of the actual heat exchange capacity of the heat exchanger to the theoretical maximum heat exchange capacity. The higher the efficiency, the better the heat exchange performance of the heat exchanger.

[0099] According to the heat exchange system heat load historical data and heat exchange system operation data, the heat exchanger heat load prediction index value is obtained through analysis and processing.

[0100] The specific method for obtaining the heat exchanger heat load prediction index value is as follows:

[0101] A first indicator value for heat exchanger heat load prediction is obtained based on analysis of historical heat load data of the heat exchange system. The first indicator value for heat exchanger heat load prediction represents quantitative data on the degree to which the historical heat load mean value and historical heat load change rate of the heat exchange system jointly affect the heat load of the heat exchanger. The first indicator value for heat exchanger heat load prediction is obtained as follows: the historical heat load mean value and reference historical heat load change rate of a reference heat exchange system stored in a database are extracted.

[0102] Normalize the historical heat load mean in the heat exchange system's heat load historical data by dividing the historical heat load mean by the reference historical heat load mean to obtain the normalized value of the historical heat load mean. Perform the same operation on the historical heat load change rate by dividing the historical heat load change rate by the reference historical heat load change rate to obtain the normalized value of the historical heat load change rate. Extract the pre-set historical heat load mean weights and historical heat load change rate weights from the database. Multiply the normalized value of the historical heat load mean by the historical heat load mean weight, and multiply the normalized value of the historical heat load change rate by the historical heat load change rate weight. Finally, add the two product results, add 1 to the sum, and input it into a logarithmic function with base e. The logarithmic function calculation result is used as the first indicator value for heat exchanger heat load prediction.

[0103] In a specific embodiment, the first indicator value for heat exchanger heat load prediction is expressed as follows:

[0104]

[0105] Among them, F is the first index value of heat exchanger heat load prediction, is the historical heat load mean of the heat exchange system, λ is the historical heat load change rate of the heat exchange system, is the reference historical heat load mean, λ0 is the reference historical heat load change rate, z3 is the historical heat load mean weight, and z4 is the historical heat load change rate weight.

[0106] It should be noted that the historical heat load mean weight and the historical heat load change rate weight are pre-set in the database, and the value range is between 0 and 1. The extraction method is, for example, to construct a mapping set with the historical heat load mean and the historical heat load change rate respectively and the historical heat load mean weight and the historical heat load change rate weight. When used, the real-time historical heat load mean and the historical heat load change rate are input into the mapping set to extract the historical heat load mean weight and the historical heat load change rate weight.

[0107] The second index value for heat exchanger heat load prediction is obtained based on the analysis of the heat exchange system operation data. The second index value for heat exchanger heat load prediction represents a quantitative result of the degree of influence of the average temperature difference between the inlet and outlet of the heat exchanger, the average pressure difference between the inlet and outlet of the heat exchanger, and the heat exchange efficiency of the heat exchanger on the operating status of the heat exchange system.

[0108] The second indicator value of heat exchanger heat load prediction is obtained as follows:

[0109] The reference heat exchanger inlet and outlet temperature difference average, the reference heat exchanger inlet and outlet pressure difference average, and the reference heat exchanger heat exchange efficiency stored in the database are extracted.

[0110] Extract the heat exchanger inlet and outlet temperature difference mean weight, heat exchanger inlet and outlet pressure difference mean weight and heat exchange efficiency weight of the heat exchanger preset in the database.

[0111]

[0112] Among them, G is the second index value of heat exchanger heat load prediction, f is the average temperature difference between the inlet and outlet of the heat exchanger, g is the average pressure difference between the inlet and outlet of the heat exchanger, h is the heat exchange efficiency of the heat exchanger, f0 is the average temperature difference between the inlet and outlet of the reference heat exchanger, g0 is the average pressure difference between the inlet and outlet of the reference heat exchanger, h0 is the heat exchange efficiency of the reference heat exchanger, is the mean weight of the temperature difference between the inlet and outlet of the heat exchanger, is the mean weight of the heat exchanger inlet and outlet pressure difference, is the heat exchange efficiency weight of the heat exchanger, and e is a natural constant.

[0113] It is important to understand that the average inlet and outlet temperature difference of a heat exchanger directly reflects the temperature transfer between the hot and cold fluids during the heat exchange process. The larger the average temperature difference, the stronger the heat transfer driving force of the heat exchanger and the higher the heat exchange efficiency. The average inlet and outlet pressure difference reflects the resistance encountered by the fluid when flowing inside the heat exchanger. The larger the average pressure difference, the greater the resistance to fluid flow, which may lead to a decrease in fluid flow rate, thereby affecting the heat exchange efficiency. The heat exchange efficiency is jointly affected by the average inlet and outlet temperature difference and the average inlet and outlet pressure difference. When other conditions remain unchanged, an increase in the average inlet and outlet temperature difference will generally increase the heat exchange efficiency, while an increase in the average inlet and outlet pressure difference may reduce the heat exchange efficiency.

[0114] It should be noted that the weight of the average temperature difference between the inlet and outlet of the heat exchanger, the weight of the average pressure difference between the inlet and outlet of the heat exchanger, and the weight of the heat exchange efficiency of the heat exchanger all have a value range of 0 to 1. When used, the preset values ​​can be directly extracted from the database. For example, the extraction method is to construct a mapping set with the average temperature difference between the inlet and outlet of the heat exchanger, the average pressure difference between the inlet and outlet of the heat exchanger, and the heat exchange efficiency of the heat exchanger and the corresponding average temperature difference weight between the inlet and outlet of the heat exchanger, the average pressure difference between the inlet and outlet of the heat exchanger, and the heat exchange efficiency weight. When used, the average temperature difference between the inlet and outlet of the heat exchanger, the average pressure difference between the inlet and outlet of the heat exchanger, and the heat exchange efficiency of the heat exchanger are input into the mapping set to extract the corresponding average temperature difference weight between the inlet and outlet of the heat exchanger, the average pressure difference between the inlet and outlet of the heat exchanger, and the heat exchange efficiency of the heat exchanger.

[0115] The first heat exchanger thermal load prediction index value and the second heat exchanger thermal load prediction index value are multiplied by the corresponding first heat exchanger thermal load prediction index weight and the second heat exchanger thermal load prediction index weight respectively, and then the products are added to obtain the heat exchanger thermal load prediction index value.

[0116] In a specific embodiment, the heat exchanger heat load prediction index value is specifically expressed as follows:

[0117] H=F*ω1+G*ω2,

[0118] Among them, H is the heat exchanger heat load prediction index value, F is the heat exchanger heat load prediction first index value, G is the heat exchanger heat load prediction second index value, ω1 is the heat exchanger heat load prediction first index weight, ω2 is the heat exchanger heat load prediction second index weight.

[0119] It should be noted that the value range of the first indicator weight of the heat exchanger thermal load prediction and the second indicator weight of the heat exchanger thermal load prediction are both 0-1. When used, the pre-set values ​​can be directly extracted from the database. The extraction method is, for example, to construct a mapping set with the first indicator value of the heat exchanger thermal load prediction and the second indicator value of the heat exchanger thermal load prediction and the corresponding first indicator weight of the heat exchanger thermal load prediction and the second indicator weight of the heat exchanger thermal load prediction respectively. When used, the first indicator value of the heat exchanger thermal load prediction and the second indicator value of the heat exchanger thermal load prediction are input into the mapping set to obtain the corresponding weights.

[0120] The heat load prediction value of the area to which the heat exchanger belongs is extracted according to the heat load prediction index value of the heat exchanger, thereby completing the heat load prediction of the heat exchange system.

[0121] It should be noted that the heat load prediction value of the area to which the heat exchanger belongs is extracted based on the heat load prediction index value of the heat exchanger. The extraction process is as follows: extract the heat load prediction values ​​corresponding to the heat load prediction index value intervals stored in the database, and map the heat load prediction values ​​corresponding to the intervals in which the heat load prediction index value of the heat exchanger is located, and record them as the heat load prediction values ​​of the area to which the heat exchanger belongs.

[0122] S4, dynamically adjusting the heat exchange system according to the heat load prediction result of the heat exchange system.

[0123] In this embodiment, the heat exchange system is dynamically adjusted according to the heat load prediction result of the heat exchange system. The specific analysis method is as follows:

[0124] Obtain meteorological data for the area where the heat exchange system is located, including the average temperature and average humidity in the area where the heat exchange system is located.

[0125] It should be noted that the meteorological data of the area to which the heat exchange system belongs is the meteorological data of the target period of heat load prediction of the area to which the heat exchange system belongs, and can be directly collected from the geographic information system.

[0126] The meteorological characterization coefficient of the area to which the heat exchange system belongs is obtained by analyzing and processing the meteorological data of the area to which the heat exchange system belongs.

[0127] The meteorological characterization coefficient of the area to which the heat exchange system belongs is obtained by extracting the reference temperature mean and the reference humidity mean stored in the database.

[0128] Normalize the mean temperature in the meteorological data for the heat exchange system's area by dividing it by the reference mean temperature to obtain a normalized mean temperature value. Perform the same process for the mean humidity by dividing it by the reference mean humidity to obtain a normalized mean humidity value. Extract the preset mean temperature and mean humidity weights from the database. Multiply the normalized mean temperature value by the mean temperature weight, and the normalized mean humidity value by the mean humidity weight. Add these products, add 1 to the sum, and input it into a base-10 logarithmic function. The output of this function serves as the meteorological characterization coefficient for the heat exchange system's area.

[0129] The meteorological characterization coefficient of the area to which the heat exchange system belongs is expressed as follows:

[0130]

[0131] Among them, K is the meteorological characterization coefficient of the area to which the heat exchange system belongs, p is the mean temperature of the area to which the heat exchange system belongs, q is the mean humidity of the area to which the heat exchange system belongs, p0 is the reference temperature mean, q0 is the reference humidity mean, z5 is the temperature mean weight, and z6 is the humidity mean weight.

[0132] It should be understood that the value range of the temperature mean weight and the humidity mean weight are both 0-1. When used, the pre-set values ​​can be directly extracted. For example, the extraction method is to construct corresponding mapping sets for the temperature mean and humidity mean and the temperature mean weight and humidity mean weight respectively. When used, the collected temperature mean and humidity mean are input into the mapping set one by one to extract the temperature mean weight and humidity mean weight.

[0133] The ideal heat load of the area to which the heat exchange system belongs is extracted based on the meteorological characterization coefficient of the area to which the heat exchange system belongs. The extraction method is as follows: the ideal heat load corresponding to each meteorological characterization coefficient interval in the database is extracted, and the ideal heat load corresponding to the interval of the meteorological characterization coefficient of the area to which the heat exchange system belongs is mapped and recorded as the ideal heat load of the area to which the heat exchange system belongs.

[0134] The heat load prediction value of the heat exchanger area is subtracted from the ideal heat load of the heat exchange system area to obtain the heat load prediction deviation value of the heat exchange system area.

[0135] The heat load prediction deviation value of the area to which the heat exchange system belongs may be greater than zero, less than zero, or equal to zero.

[0136] The heat load prediction deviation value of the area to which the heat exchange system belongs is used as the heat load prediction result of the heat exchange system.

[0137] The heat exchanger adjustment parameters are extracted according to the heat load prediction deviation value of the area to which the heat exchange system belongs.

[0138] It should be noted that when the heat load prediction deviation value of the area to which the heat exchange system belongs is greater than zero, it means that the current predicted heat load exceeds the ideal heat load of the area. In order to prevent waste of resources, the corresponding extracted heat exchanger adjustment parameter should be less than zero, and the larger the heat load prediction deviation value of the area to which the heat exchange system belongs, the higher the degree of exceeding the ideal heat load, and the larger the absolute value of the heat exchanger adjustment parameter should be. When the heat load prediction deviation value of the area to which the heat exchange system belongs is less than zero, it means that the current predicted heat load has not reached the ideal level and the predicted heat load is less. The corresponding extracted heat exchanger adjustment parameter should be greater than zero, and the larger the absolute value of the heat load prediction deviation value of the area to which the heat exchange system belongs, the more obvious the gap between the heat load and the ideal value, and the larger the heat exchanger adjustment parameter should be.

[0139] The heat exchanger adjustment parameters include the expansion valve opening adjustment value and the variable frequency pump speed.

[0140] Extract the initial parameters of the heat exchanger.

[0141] The heat exchange system is dynamically adjusted according to the initial parameters and adjustment parameters of the heat exchanger.

[0142] In a specific embodiment, it is assumed that the initial expansion valve opening of the heat exchanger is 50% and the variable frequency pump speed is 1000 rpm. After calculation, it is found that the heat load prediction deviation value of the area to which the heat exchange system belongs is greater than zero and is 10kW. According to the rules, the corresponding expansion valve opening adjustment value extracted is -10%, and the variable frequency pump speed adjustment value is -200 rpm. Then, based on the initial parameters of the heat exchanger and these adjustment parameters, dynamic adjustment is performed, and the adjusted expansion valve opening becomes 50% + (-10%) = 40%, and the variable frequency pump speed becomes 1000 + (-200) = 800 rpm. Through dynamic adjustment, the refrigerant flow and circulation power are reduced, avoiding resource waste when the heat load prediction value is higher than the ideal value, and ensuring that the heat exchange system can operate more efficiently and reasonably.

[0143] See Figure 2 As shown, the second aspect of the present invention provides a heat exchange system optimization system based on big data analysis, comprising:

[0144] The computing resource demand assessment module is used to record the heat load prediction data streams of the heat exchange system as data streams, perform computing resource demand assessment on each data stream, analyze the computing resource demand assessment indicators of each data stream, and thus perform resource restrictions.

[0145] The sampling frequency adjustment module is used to generate adaptive sampling frequency control instructions based on the computing resource demand evaluation indicators of each data stream, and synchronously analyze the real-time data flow fluctuation coefficient and system computing power evaluation indicators to adjust the sampling frequency of each data stream.

[0146] The heat load prediction module is used to collect data based on the adjusted sampling frequency, thereby predicting the heat load of the heat exchange system.

[0147] The heat exchange system dynamic adjustment module is used to dynamically adjust the heat exchange system according to the heat load prediction results of the heat exchange system.

[0148] It will be understood by those skilled in the art that embodiments of the present invention may be provided as methods, systems, or computer program products. Thus, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware. Furthermore, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0149] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowcharts and / or block diagrams, as well as combinations of processes and / or blocks in the flowcharts and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowcharts and / or block diagrams. Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0150] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.

[0151] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing the instructions executed on the computer or other programmable device for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1 The steps for the function specified in one or more boxes.

[0152] Although the preferred embodiments of the present invention have been described, those skilled in the art may make additional changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the present invention.

[0153] Obviously, those skilled in the art may make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if such changes and modifications fall within the scope of the claims and their equivalents, the present invention is intended to include such changes and modifications.

Claims

1. A heat exchange system optimization method based on big data analysis, characterized in that: The following steps are involved: S1, record each heat load prediction data stream of the heat exchange system as each data stream, perform computing resource demand assessment on each data stream, analyze computing resource demand assessment indicators of each data stream, and thus perform resource restriction; S2: Based on the computing resource demand evaluation indicators of each data stream, and synchronously analyzing the real-time data flow fluctuation coefficient and the system computing power evaluation indicators, an adaptive sampling frequency control instruction is generated to adjust the sampling frequency of each data stream. The specific analysis process of adjusting the sampling frequency of each data stream is as follows: Extracting a first correction factor for the sampling frequency of each data stream according to the computing resource demand evaluation index of each data stream; the larger the computing resource demand evaluation index, the smaller the first correction factor for the sampling frequency; Extract the second correction factor of the sampling frequency based on the system computing power evaluation index; The larger the system computing power evaluation index is, the larger the second correction factor of the sampling frequency is; According to the real-time data flow fluctuation coefficient of each data flow, the first correction factor of the sampling frequency of each data flow, and the second correction factor of the sampling frequency of each data flow, a sampling adjustment index of each data flow is obtained through comprehensive analysis and processing; Extracting the preset sampling adjustment indicator threshold in the database; Subtracting the sampling adjustment index threshold from the sampling adjustment index of each data stream to obtain the sampling adjustment deviation index of each data stream; Extracting a sampling frequency adjustment value for each data stream according to a sampling adjustment deviation indicator for each data stream; Adjusting the sampling frequency of each data stream according to the initial sampling frequency of each data stream and the sampling frequency adjustment value of each data stream; S3, collecting data based on the adjusted sampling frequency, thereby predicting the heat load of the heat exchange system; S4, dynamically adjusting the heat exchange system according to the heat load prediction result of the heat exchange system.

2. The heat exchange system optimization method based on big data analysis according to claim 1, characterized in that: The heat exchange system heat load prediction data streams specifically include: heat exchange system heat load historical data, heat exchange system operation data and meteorological data of the area to which the heat exchange system belongs.

3. The heat exchange system optimization method based on big data analysis according to claim 1, characterized in that: The computing resource requirement evaluation indicators of each data flow are analyzed to impose resource restrictions. The specific process is as follows: Obtain historical resource parameters for each data stream and analyze the computing resource demand assessment indicators for each data stream; Extract the demand quantities corresponding to the demand evaluation index intervals stored in the database, and map and extract the demand quantities corresponding to the intervals in which the computing resource demand evaluation indexes of each data flow are located, and record them as the computing resource demand quantities of each data flow; Resource restrictions are imposed based on the computing resource requirements of each data flow.

4. The heat exchange system optimization method based on big data analysis according to claim 3, characterized in that: The specific analysis process of analyzing the computing resource demand evaluation indicators of each data flow is as follows: Obtain historical resource parameters for each data stream, including the historical average data volume, historical average data processing frequency, historical average number of data types, and historical average data transmission bandwidth of each data stream; Get reference history resource parameters; Based on the historical resource parameters of each data flow and the reference historical resource parameters, a comprehensive analysis is performed to obtain the computing resource demand assessment indicators of each data flow; The computing resource demand evaluation index of each data flow represents the quantitative result of the impact of the historical data volume average, historical data processing frequency average, historical data type number average and historical data transmission bandwidth average of each data flow on the degree of computing resource demand of the data flow.

5. The heat exchange system optimization method based on big data analysis according to claim 1, characterized in that: The specific analysis process of the real-time data traffic fluctuation coefficient is as follows: Extract the initial sampling frequency and data collection verification duration of each data stream; During the data collection and verification period, data of each data stream is collected at the initial sampling frequency of each data stream, thereby obtaining the data fluctuation parameters of each data stream; The data fluctuation parameters of each data stream include the average data packet arrival interval, the total number of data packets, the peak data rate and the average data rate of each data stream; The real-time data flow fluctuation coefficient of each data flow is obtained based on the data fluctuation parameter analysis of each data flow; The real-time data flow fluctuation coefficient of each data flow represents a quantitative result of the degree of influence of the average arrival interval of data packets, the total number of data packets, the peak data rate and the average data rate of each data flow on the data flow fluctuation state.

6. The heat exchange system optimization method based on big data analysis according to claim 1, characterized in that: The specific analysis method for the system computing power evaluation indicator is as follows: Extract the system's CPU idle rate, remaining memory, and GPU idle rate from the system program log and analyze them to obtain the system computing power evaluation indicators; The system computing power evaluation index represents the quantitative result of the degree to which the system's CPU idle rate, memory remaining amount, and GPU idle rate jointly affect the system's remaining computing power status.

7. The heat exchange system optimization method based on big data analysis according to claim 1, characterized in that: The specific analysis method for predicting the heat load of the heat exchange system is as follows: During the preset monitoring period, data is collected based on the adjusted sampling frequency to obtain the historical heat load data and operation data of the heat exchange system; The heat load history data of the heat exchange system includes the historical heat load mean value and the historical heat load change rate of the heat exchange system; The operating data of the heat exchange system includes the average temperature difference between the inlet and outlet of the heat exchanger, the average pressure difference between the inlet and outlet of the heat exchanger, and the heat exchange efficiency of the heat exchanger; According to the heat exchange system heat load historical data and heat exchange system operation data, the heat exchanger heat load prediction index value is obtained by analysis and processing; The heat load prediction value of the area to which the heat exchanger belongs is extracted according to the heat load prediction index value of the heat exchanger, thereby completing the heat load prediction of the heat exchange system.

8. The heat exchange system optimization method based on big data analysis according to claim 1, characterized in that: The dynamic adjustment of the heat exchange system is performed according to the heat load prediction result of the heat exchange system. The specific analysis method is as follows: Obtain meteorological data of the area where the heat exchange system is located, including the average temperature and average humidity of the area where the heat exchange system is located; The meteorological characterization coefficient of the area to which the heat exchange system belongs is obtained by analyzing and processing the meteorological data of the area to which the heat exchange system belongs; Extract the ideal heat load of the area where the heat exchange system belongs according to the meteorological characterization coefficient of the area where the heat exchange system belongs; The heat load prediction value of the heat exchanger area is subtracted from the ideal heat load of the heat exchange system area to obtain the heat load prediction deviation value of the heat exchange system area; The heat load prediction deviation value of the area to which the heat exchange system belongs is used as the heat load prediction result of the heat exchange system; Extracting heat exchanger adjustment parameters based on the heat load prediction deviation value of the area to which the heat exchange system belongs; The heat exchanger adjustment parameters include the expansion valve opening adjustment value and the variable frequency pump speed; Extract the initial parameters of the heat exchanger; The heat exchange system is dynamically adjusted according to the initial parameters and adjustment parameters of the heat exchanger.

9. A system using the heat exchange system optimization method based on big data analysis as described in any one of claims 1 to 8, characterized in that: include: A computing resource demand assessment module is used to record each heat load prediction data stream of the heat exchange system as each data stream, perform computing resource demand assessment on each data stream, analyze the computing resource demand assessment index of each data stream, and thereby perform resource restriction; The sampling frequency adjustment module is used to generate adaptive sampling frequency control instructions based on the computing resource demand evaluation indicators of each data stream, and simultaneously analyze the real-time data flow fluctuation coefficient and system computing power evaluation indicators to adjust the sampling frequency of each data stream; A heat load prediction module is used to collect data based on the adjusted sampling frequency, thereby predicting the heat load of the heat exchange system; The heat exchange system dynamic adjustment module is used to dynamically adjust the heat exchange system according to the heat load prediction results of the heat exchange system.

Images