Control Method, Device, and Storage Medium of an Integrated Scheduling High-Efficiency Refrigerating Machine Room System

Abstract

The present application discloses a control method, device and storage medium for a fusion scheduling efficient refrigeration machine room system, which relates to the technical field of central air-conditioning control. The method includes: obtaining the operation data of each edge device and the sensor data in each area; calculating the energy efficiency index of each subsystem based on the operation data, the allocation weight of the subsystem corresponding to each device type, the comfort coefficient associated with each area, and the sensor data in each area; formulating the control strategy of each subsystem according to the energy efficiency index and combining the cold quantity prediction data associated with each area; reversely decomposing the control strategy to obtain the control parameters of the edge devices in each area, and sending the control parameters to the edge devices in the corresponding area. The present application overcomes the technical defect of poor refrigeration effect of the central air-conditioning when scheduling the cold source machine room, and improves the global energy utilization rate and refrigeration effect.

Description

Technical Field

[0001] This application relates to the field of central air-conditioning control technology, and particularly to a control method, device, and storage medium for an integrated scheduling and efficient refrigerating machine room system. Background Art

[0002] As the main energy-consuming system in a building, the cold source system of a central air-conditioning system contains a cold station module composed of multiple modules, and the cold station module accounts for up to 80% of the energy consumption of the central air-conditioning system. Therefore, controlling the energy consumption of the cold source system helps to save energy and reduce emissions.

[0003] In related technologies, the cold source system generally uses a PLC (Programmable Logic Controller) or DDC (Direct Digital Control) module to construct the basic control logic, and the PLC and DDC modules rely on the PID (Proportional Integral Derivative Control Algorithm). However, the parameters of the PID algorithm are fixed and not suitable for processing non-linear data relationships, and thus it is difficult to adjust adaptively. As a result, when scheduling the cold source machine room, the refrigeration effect of the central air-conditioning system is not good.

[0004] The above content is only used to assist in understanding the technical solution of this application, and does not represent an admission that the above content is prior art. Summary of the Invention

[0005] The main purpose of this application is to provide a control method for an integrated scheduling and efficient refrigerating machine room system, aiming to solve the technical problem that the refrigeration effect of the central air-conditioning system is not good when scheduling the cold source machine room.

[0006] To achieve the above purpose, this application proposes a control method for an integrated scheduling and efficient refrigerating machine room system, the method includes: obtaining the operation data of each edge device and the sensor data in each area;

[0007] Based on the operation data, the distribution weights of the subsystems corresponding to each device type, the comfort coefficients associated with each area, and the sensor data in each area, calculate the energy efficiency indicators of each subsystem;

[0008] According to the energy efficiency indicators, combined with the cold quantity prediction data associated with each area, formulate the control strategies for each subsystem;

[0009] Decompose the control strategies in reverse to obtain the control parameters of the edge devices in each area, and send the control parameters to the edge devices in the corresponding area.

[0010] In one embodiment, the edge-side central unit receives a start instruction and parses the corresponding service information according to the start instruction;

[0011] Based on the service information, configure the logical mapping relationship between the edge-side device corresponding to the start instruction and the service;

[0012] Based on the basic parameters of the edge-side device, adjust the basic parameters in combination with the logical mapping relationship to generate the policy parameters;

[0013] Configure the policy parameters on the corresponding edge-side device and control the edge-side device to operate according to the corresponding policy parameters.

[0014] In one embodiment, obtain the original operation data of each edge-side device and the original sensor data transmitted back by the sensors in each area;

[0015] Based on the original operation data and the original sensor data, perform data preprocessing to generate the operation data and the sensor data, and store the operation data and the sensor data in a data pool.

[0016] In one embodiment, based on the sensor data, extract the basic features corresponding to the sensor data and convert the basic features into derivative features;

[0017] Based on the derivative features, build a model through the random forest algorithm to generate a comfort evaluation model;

[0018] The comfort evaluation model uses the basic features and user feedback information as training samples to optimize the comfort evaluation model through the training samples;

[0019] The comfort evaluation model calculates the comfort coefficient of the current environment according to the sensor data.

[0020] In one embodiment, based on the operation data, perform energy efficiency analysis on each subsystem to determine the energy efficiency ratio of each subsystem to the overall system;

[0021] According to the rated energy consumption of the edge-side devices in each subsystem, perform weighted fitting, and combine the allocation weights to obtain the energy efficiency priority coefficient of each subsystem;

[0022] Based on the energy efficiency priority coefficient, the sensor data, in combination with the comfort coefficient and the energy efficiency ratio, generate the energy efficiency index.

[0023] In one embodiment, based on the operation data, perform energy efficiency analysis on each subsystem to determine the energy efficiency ratio of each subsystem to the overall system;

[0024] The cooling capacity prediction model is trained and verified by using the multivariate linear regression algorithm based on historical cooling capacity demand samples to generate a cooling capacity demand prediction model;

[0025] The cooling capacity demand prediction model calculates the cooling capacity prediction data of the current environment according to the operation data.

[0026] In one embodiment, a particle swarm optimization algorithm is performed based on each of the subsystems, and several global optimal solutions are obtained by combining the energy efficiency index and the cooling capacity prediction data;

[0027] By comparing and judging each of the global optimal solutions, a target global optimal solution is obtained;

[0028] The target global optimal solution is analyzed to obtain the target local optimal solutions of the corresponding subsystems;

[0029] According to the target local optimal solutions, the control strategies of the corresponding subsystems are generated.

[0030] In one embodiment, based on the control strategies corresponding to the subsystems, the edge control strategies corresponding to the regions are generated by each of the edge - side centers;

[0031] The edge control strategies are reversely decomposed to generate the control parameters of the edge - side devices in each of the regions, and the control parameters are sent to the corresponding edge - side devices;

[0032] The edge - side devices operate based on the corresponding control parameters.

[0033] In addition, to achieve the above object, the present application also provides a device for fusing and scheduling high - efficiency refrigeration plant equipment, the device includes: a memory, a processor, and a computer program stored on the memory and executable on the processor, and the computer program is configured to implement the steps of the control method of the fusing and scheduling high - efficiency refrigeration plant system as described above.

[0034] In addition, to achieve the above object, the present application also provides a storage medium, the storage medium is a computer - readable storage medium, and a computer program is stored on the storage medium, and when the computer program is executed by a processor, the steps of the control method of the fusing and scheduling high - efficiency refrigeration plant system as described above are implemented.

[0035] The present application provides a control method for an integrated scheduling and highly efficient refrigeration machine room system, including obtaining the operation data of each edge device and the sensor data in each area; calculating the energy efficiency indicators of each subsystem based on the operation data, the distribution weights of the subsystems corresponding to each device type, the comfort coefficients associated with each area, and the sensor data in each area; formulating the control strategies of each subsystem according to the energy efficiency indicators, in combination with the cooling capacity prediction data associated with each area; reversely decomposing the control strategies to obtain the control parameters of the edge devices in each area, and sending the control parameters to the edge devices in the corresponding area. Through the integrated scheduling of each subsystem and the overall system, combined with data analysis and strategy regulation, this solution realizes the closed-loop optimization of the "perception - decision - execution" of the building energy system, and improves the continuous expansion ability of the cold source in the direction of cost reduction and efficiency improvement.

[0036] In summary, by analyzing the associated data of each area, regulating the control strategies of each subsystem, and then scheduling the control parameters of the edge devices, the present application seeks the global optimal energy efficiency without sacrificing user comfort, and even improving user comfort, overcomes the technical defect of poor refrigeration effect of the central air conditioner when scheduling the cold source machine room, and improves the global energy utilization rate and refrigeration effect. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] The accompanying drawings herein are incorporated into the specification and form a part of the specification, showing embodiments consistent with the present application, and are used together with the specification to explain the principles of the present application.

[0038] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the accompanying drawings required for use in the description of the embodiments or the prior art. Obviously, for those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0039] Figure 1 It is a schematic flowchart of the first embodiment of the control method for the integrated scheduling and highly efficient refrigeration machine room system of the present application;

[0040] Figure 2 It is a schematic flowchart of the second embodiment of the control method for the integrated scheduling and highly efficient refrigeration machine room system of the present application;

[0041] Figure 3 It is a schematic flowchart of the third embodiment of the control method for the integrated scheduling and highly efficient refrigeration machine room system of the present application;

[0042] Figure 4 It is a schematic flowchart of the fourth embodiment of the control method for the integrated scheduling and highly efficient refrigeration machine room system of the present application;

[0043] Figure 5 It is a schematic flowchart of the fifth embodiment of the control method for the integrated scheduling efficient refrigeration machine room system of the present application;

[0044] Figure 6 It is a schematic flowchart of the sixth embodiment of the control method for the integrated scheduling efficient refrigeration machine room system of the present application;

[0045] Figure 7 It is a schematic flowchart of the seventh embodiment of the control method for the integrated scheduling efficient refrigeration machine room system of the present application;

[0046] Figure 8 It is a schematic flowchart of the eighth embodiment of the control method for the integrated scheduling efficient refrigeration machine room system of the present application;

[0047] Figure 9 It is an architecture diagram of the closed-loop control of the integrated scheduling efficient refrigeration machine room system of the present application;

[0048] Figure 10 It is a flowchart of the integrated scheduling efficient refrigeration machine room system module of the present application;

[0049] Figure 11 It is a structure diagram of the edge device of the present application;

[0050] Figure 12 It is a schematic structural diagram of the integrated scheduling efficient refrigeration machine room equipment of the present application.

[0051] The realization of the purpose, functional features and advantages of the present application will be further described with reference to the embodiments and the accompanying drawings. Detailed implementation manners

[0052] It should be understood that the specific embodiments described herein are only used to explain the technical solutions of the present application and are not used to limit the present application.

[0053] In the related art, the cold source system generally uses PLC or DDC modules to build the basic control logic, and the PLC and DDC modules rely on the PID algorithm. However, the parameters of the PID algorithm are fixed and not suitable for processing non-linear data relationships, and thus it is difficult to adjust adaptively. As a result, when scheduling the cold source machine room, the refrigeration effect of the central air conditioner is not good.

[0054] The present application provides a solution: First, obtain the operation data of each edge device and the sensor data in each area. Then, based on the operation data, the allocation weights of the subsystems corresponding to each device type, the comfort coefficients associated with each area, and the sensor data in each area, calculate the energy efficiency indicators of each subsystem. Next, according to the energy efficiency indicators, combined with the cooling capacity prediction data associated with each area, formulate the control strategies for each subsystem. Finally, decompose the control strategies in reverse to obtain the control parameters of the edge devices in each area, and send the control parameters to the edge devices in the corresponding area. This overcomes the technical defect of poor cooling effect of the central air conditioner when scheduling the cold source machine room, and improves the global energy utilization rate and cooling effect.

[0055] It should be noted that the execution subject of this embodiment can be a computing service device with data processing, network communication, and program running functions, such as a tablet computer, a personal computer, a mobile phone, etc., or an electronic device, an integrated scheduling high-efficiency refrigeration machine room device, an integrated scheduling high-efficiency refrigeration machine room system, etc. that can implement the above functions. Hereinafter, taking the integrated scheduling high-efficiency refrigeration machine room system as an example, this embodiment and the following embodiments will be described.

[0056] To better understand the technical solution of the present application, the following will be described in detail in combination with the specification drawings and specific implementation manners.

[0057] The embodiment of the present application provides a control method for an integrated scheduling high-efficiency refrigeration machine room system, referring to Figure 1 , Figure 1 is a schematic flowchart of the first embodiment of the control method for the integrated scheduling high-efficiency refrigeration machine room system of the present application.

[0058] In this embodiment, the control method for the integrated scheduling high-efficiency refrigeration machine room system includes steps S10 to S40:

[0059] Step S10, obtain the operation data of each edge device and the sensor data in each area.

[0060] In this embodiment, each area is a single air-conditioning operation system, including an edge central unit and the edge devices covered by the edge central unit. The edge device is an electromechanical device that maintains the operation of the air conditioner in each area. The operation data is various parameter data of the edge device during operation. The sensor data is the environmental data received by the sensors in the area.

[0061] As an alternative implementation, the edge-side center collects the original operation data of each edge-side device and the original sensor data related to each region, and performs preprocessing operations such as data cleaning, data supplementation, and data missing supplementation on the collected original operation data and sensor data to obtain operation data and sensor data.

[0062] Step S20: Calculate the energy efficiency index of each subsystem based on the operation data, the distribution weights corresponding to each device type subsystem, the comfort coefficients associated with each region, and the sensor data within each region.

[0063] In this embodiment, a subsystem is composed of a class of edge-side devices of the same type or the same function, such as a cooling subsystem, a refrigeration subsystem, and a chiller. The distribution weight of the subsystem is a dynamic adjustment coefficient set for each subsystem based on business priority or energy consumption ratio, which is used to quantify its impact on the overall energy efficiency. In the calculation of the energy efficiency index, the energy consumption of each subsystem is allocated through the distribution weight. The comfort coefficient is a quantified index calculated through environmental parameters such as temperature, humidity, and air velocity, which reflects the threshold range of the user's satisfaction with the regional environment. The energy efficiency index is a comprehensive evaluation parameter generated by combining energy consumption data, equipment efficiency, and environmental requirements, which is used to measure the effective output per unit energy consumption of the subsystem.

[0064] As an alternative implementation, according to the operation data of each edge-side device, the energy efficiency of edge-side devices of the same type is summed and compared with the energy efficiency of the overall system to obtain the energy efficiency ratio of each subsystem. The sensor data obtains the current comfort coefficient through a preset comfort calculation rule and compares it with the comfort coefficient associated with the region. Taking the current comfort coefficient not being lower than the comfort coefficient associated with the region as the bottom line, and then weighted fitting based on the rated energy consumption of the edge-side devices within each subsystem, this is used as the upper limit of the operation energy consumption of the edge-side devices. Finally, according to the energy efficiency ratio and the distribution weights of each subsystem, the energy efficiency index is generated according to a preset energy efficiency index rule, such as energy efficiency index = energy efficiency ratio of each subsystem × distribution weight - (current comfort coefficient - comfort coefficient associated with the region).

[0065] Step S30: Develop a control strategy for each subsystem according to the energy efficiency index and in combination with the cooling capacity prediction data associated with each region.

[0066] In this embodiment, the cooling capacity prediction data is a predicted value of the future cooling demand of the region generated based on historical operation data and a machine learning model. The control strategy is an equipment operation rule formulated according to the energy efficiency target and the cooling capacity demand, which realizes the minimization of energy consumption under the constraint of ensuring comfort.

[0067] As an alternative implementation, based on the minimum standard of energy efficiency indicators, the operation of the subsystem is adjusted by combining the cooling capacity prediction data. The cooling subsystem, the refrigeration subsystem, and the chiller are regarded as three particle swarms through the particle swarm optimization algorithm, and their global optimal solutions are calculated respectively. For example, reducing the cooling tower fan speed by 20% + adjusting the frequency of the chilled water pump to 45 Hz can save 18% of electricity and meet the cooling capacity demand. By comparison, the target global optimal solution is determined, and the control strategies for each subsystem are generated based on the target global optimal solution.

[0068] Step S40: Reverse decompose the control strategy to obtain the control parameters of the edge devices in each region, and send the control parameters to the edge devices in the corresponding region.

[0069] In this embodiment, reverse decomposition disassembles the global control strategy into device-level executable instructions through mathematical inversion or a rule engine. The control parameters of the edge devices are the parameter values that drive the devices to operate and need to meet the boundary conditions for the safe operation of the devices.

[0070] As an alternative implementation, the control strategies of each subsystem are disassembled through a distributed optimization algorithm, and then an edge control strategy is generated by the edge center. According to the edge control strategy, the control parameters of the edge devices in each region are reverse decomposed, and the control parameters are sent to the edge devices. The edge devices detect the adaptability of the control parameters. If the adaptability test passes, the edge devices operate according to the corresponding control parameters. If the adaptability test fails, an alarm is uploaded and the "guaranteed parameters" are run.

[0071] Exemplarily, the air conditioning system collects the operation data such as the current of the air conditioning host and the wind speed gear installed on each floor in real time through the edge center, combines the temperature and humidity sensors and the infrared passenger flow counter to monitor the environmental status of each region, and uses the edge computing node to correlate the device energy consumption with the room temperature change to generate an analysis index of "power consumption - temperature deviation". According to the comfort requirements of the meeting rooms and shops and the cooling capacity prediction curve for the next two hours, the system automatically generates strategies such as reducing the frequency of the chilled water pump by 15% and starting and stopping the fresh air units at intervals through an optimization algorithm. Finally, these instructions are converted into the temperature setting values (26 ± 0.3 °C) and valve opening parameters (55% - 70%) of specific devices and sent to the corresponding area controllers in real time through the Internet of Things platform for execution. While dynamically maintaining the indoor environment, the energy consumption is reduced by 18% compared with the traditional mode, and the filter cleaning warning is triggered 2 times.

[0072] By analyzing the associated data of each region, regulating the control strategies of each subsystem, and then scheduling the control parameters of the edge devices, the global optimal energy efficiency is sought without sacrificing user comfort, or even improving user comfort, overcoming the technical defect of poor cooling effect of the central air conditioner when scheduling the cold source machine room, and improving the global energy utilization rate and cooling effect.

[0073] Based on any of the above embodiments, in the second embodiment of the present application, refer to Figure 2 , Figure 2 which is a schematic flowchart of the second embodiment of the control method for the integrated scheduling efficient refrigeration machine room system of the present application. Before the step S10, the steps A11 to A14 are further included:

[0074] Step A11, the edge - side center receives a start instruction and parses the corresponding service information according to the start instruction.

[0075] In this embodiment, the edge - side center is an intelligent decision - making node deployed at the device end, responsible for receiving and parsing cloud instructions, coordinating local devices to execute tasks, and assisting the cloud in collecting various data. The start instruction is a structured command defined in a lightweight data exchange format for the user to turn on the air conditioner in the area, and it contains service information. The service information is the information of the command issued by the user and contains the functions to be run.

[0076] As an alternative implementation, when the edge - side center receives the start instruction from the user, it extracts the service information in the start instruction through the built - in protocol parsing module and verifies the data integrity. If the data is complete, it configures the logical mapping relationship corresponding to the service information. If the data is incomplete, it configures according to the preset logical mapping relationship.

[0077] Step A12, based on the service information, configure the logical mapping relationship between the edge - side device corresponding to the start instruction and the service.

[0078] In this embodiment, the logical mapping relationship between the edge - side device and the service is to map the parsed instruction to specific edge - side device actions according to a preset rule library or script framework, define the association rules between device parameters and service requirements, and achieve dynamic parameter adaptation through conditional judgment or function mapping.

[0079] As an alternative implementation, based on the service information, the edge - side center will automatically perform dynamic binding of each service instruction to a list of edge - side devices that meet the conditions. The specific process is as follows: Parse the trigger conditions in the instruction through the business rule engine and match the device attribute library, such as device ID, installation location, and function type, to generate the logical mapping relationship of "service - device".

[0080] Step A13, based on the basic parameters of the edge - side device, combine with the logical mapping relationship to adjust the basic parameters and generate the policy parameters.

[0081] In this embodiment, the basic parameters of the edge device are the hardware performance and operating configuration preset at the factory of the device, which constitute the device function baseline. The policy parameter is the device operation control value generated through optimization and is used to guide the device to execute specific business policies.

[0082] As an alternative implementation, the edge central device will first read the device basic parameters, combined with the logical mapping relationship. For example, when executing the business of cooling to 24 degrees Celsius, the cold air at 24 degrees Celsius will last for 2 minutes, etc. The conditional logic is converted into a mathematical constraint through a rule engine or script parsing, and the policy parameters that the device can execute are dynamically calculated. For example, the power setting value range = (rated upper limit × 0.8, current load × 1.2).

[0083] Step A14: Configure the policy parameters on the corresponding edge device, and control the edge device to operate according to the corresponding policy parameters.

[0084] In this embodiment, the edge device operates based on the corresponding policy parameters under the guidance of the policy parameters.

[0085] As an alternative implementation, by writing each policy parameter into each edge device, the operation mechanism of the edge device is generated according to the policy parameters, so that the edge device operates according to the operation mechanism and detects the operation state, and the operation state is fed back to the edge central device in real time.

[0086] Exemplarily, in a central air-conditioning system, the edge thermostat is initially set to a fixed output of 24°C in the cooling mode, the fan is in the medium speed gear, and the basic frequency of the compressor is 40 Hz. Based on the preset "temperature difference - adjustment coefficient" logical mapping, combined with the real-time collected indoor carbon dioxide concentration, heat imaging data of the number of people, and the perceived temperature forecast released by the meteorological station, the optimization policy parameters are dynamically generated through a fuzzy control algorithm, and finally the temperature fluctuation in the office area is controlled within ±0.3°C.

[0087] Due to the regulation of the edge central device, the air conditioners in each area can operate independently and optimize the policy, improving the operation efficiency of the system.

[0088] Based on any of the above embodiments, in the third embodiment of the present application, refer to Figure 3 , Figure 3 This is the schematic flowchart of the control method for the third embodiment of the integrated scheduling high-efficiency refrigeration machine room system of the present application. The step S10 includes steps B11 to B12:

[0089] Step B11: Obtain the original operation data of each edge device and the original sensor data transmitted back by the sensors in each area.

[0090] In this embodiment, the original operation data is the initial data directly collected by the edge device during operation, without filtering, processing, or aggregation. Sensor original data refers to the unprocessed environmental data directly obtained by sensors.

[0091] As an alternative implementation, the original operation data of each edge device and the original sensor data within each region are obtained through the edge center set in each region. The obtained sensor original data is associated with the obtained region to obtain the sensor original data of that region.

[0092] Step B12: Based on the original operation data and the original sensor data, perform data preprocessing to generate the operation data and the sensor data, and store the operation data and the sensor data in the data pool.

[0093] In this embodiment, data preprocessing is an operation process of cleaning the basic data, removing sensor transient noise, normalizing, filling missing values, and linearly interpolating to complete breakpoint data.

[0094] As an alternative implementation, perform data cleaning, normalization, and missing value filling on the original operation data and the original sensor data to generate the operation data of the edge device and the sensor data of the associated region, and store the operation data and the sensor data in the data pool.

[0095] Exemplarily, the central air-conditioning system automatically collects the operation data such as the power consumption and wind speed of each device through the edge center in each region, and at the same time connects temperature and humidity sensors to count the real-time temperature and the number of people in each region. The system first cleans the data to remove error values, then calculates the average energy consumption and the room temperature change trend every 10 minutes to generate analysis results such as "peak power consumption period" and "optimal temperature setting", and finally encrypts and stores these data in the database. When the number of people in a certain region increases greatly, the system automatically lowers the air-conditioning temperature and reduces the air volume in the idle area to achieve overall power saving while keeping the indoor temperature stable.

[0096] Since the operation data and sensor data of each region are automatically collected and stored for analysis, it provides a data basis for the prediction of control strategies and fusion scheduling, and improves the response efficiency of the system.

[0097] Based on any of the above embodiments, in the fourth embodiment of the present application, refer to Figure 4 , Figure 4 This is a schematic flowchart of the fourth embodiment of the control method for the fusion scheduling high-efficiency refrigeration machine room system of the present application. Before step S20, steps C11 to C14 are further included:

[0098] Step C11: Based on the sensor data, extract the basic features corresponding to the sensor data, and convert the basic features into derivative features.

[0099] In this embodiment, the basic features are calculated from the corresponding environmental indicators, such as the maximum value, average value, and weighted value of the environmental indicators, which reflect the surface characteristics of the data. The derivative features are composite indicators generated through mathematical operations or domain knowledge.

[0100] As an alternative implementation, the environmental indicators in the sensor data are extracted through the edge-side central unit. The environmental indicators are the basic features, such as temperature, humidity, and wind speed. The basic features are then subjected to more complex calculations, such as analyzing the data fluctuation pattern and periodic change (for example, equivalent temperature (Teq = Ta - 0.4(RH - 50)), wind chill index (WCI = 13.12 + 0.6215Ta - 11.37v0.16 + 0.3965Tav0.16), temperature-humidity entropy (S = Ta × RH / 100), where Ta represents air temperature, RH represents relative humidity expressed as a percentage, v represents wind speed, and Tav represents the interaction term between Ta and v, which means multiplying Ta and v. It is necessary to ensure that the wind speed unit matches the formula) to generate derivative features.

[0101] Step C12: Based on the derivative features, build a comfort evaluation model through the random forest algorithm.

[0102] In this embodiment, the random forest algorithm is a machine learning method based on the integration of multiple decision trees. It improves the robustness of the model through a voting mechanism or average prediction and is suitable for modeling high-dimensional features and non-linear relationships. The comfort evaluation model evaluates comfort through sensor data and user feedback information.

[0103] As an alternative implementation, first divide the standardized derivative feature dataset into a training set and a validation set according to the timestamp, input it into the random forest algorithm for multi-tree parallel training, evaluate the feature importance through the Gini coefficient and optimize the node splitting rule. After training, use cross-validation to calculate the accuracy of the comfort model. If the accuracy is greater than 90%, it is set as the comfort evaluation model. If the accuracy does not meet the standard, continue training.

[0104] Step C13: The comfort evaluation model uses the basic features and user feedback information as training samples to optimize the comfort evaluation model through the training samples.

[0105] In this embodiment, user feedback information is completed by users uploading evaluations and ratings on their own initiative. The training samples are paired data sets containing sensor data and user feedback, which are used for learning and optimization of the comfort evaluation model.

[0106] As an optional implementation, the basic features are aligned with the user feedback information by timestamp, a labeled training sample set is generated through sliding window aggregation, noise data is removed, and the comfort assessment model is optimized and trained using the training sample set.

[0107] Step C14: the comfort evaluation model calculates the comfort coefficient of the current environment according to the sensor data.

[0108] In this embodiment, the comfort coefficient is a numerical indicator that quantifies the comfort level of the environment and can be mapped to a graded label or a recommended control action.

[0109] As an optional implementation method, based on the environmental data collected by sensors in real time, such as temperature, humidity, PM2.5, and light intensity, the data is first cleaned and standardized, such as eliminating abnormal mutation values ​​and unifying units to obtain sensor data. These features are input into the pre-trained random forest model, and the current derived features are weighted and summed through multiple decision trees, such as comfort coefficient = temperature × temperature weight + humidity × humidity weight + wind speed × wind speed weight, and the comfort coefficient of 0 to 100 points is output.

[0110] For example, the central air-conditioning system deploys temperature and humidity sensors, carbon dioxide concentration detection modules and fan status monitoring units on each floor to collect real-time operating data such as equipment current and wind speed, extract basic features through the edge center, and generate derived features based on the formula. The comfort model is trained by integrating historical user APP scores through the random forest algorithm. When a sudden increase in carbon dioxide in the conference room is detected, the comfort assessment model triggers the linkage control of increasing the air volume and adjusting the cold water valve opening to 65%, and sends the control parameters to the corresponding edge device.

[0111] Due to the training and application of the comfort assessment model, the system is able to reduce costs and increase efficiency of cold source rooms through the integration of AI and other technologies such as the Internet of Things without reducing user comfort, and enhance the continuous expansion capabilities of cold sources in the direction of reducing costs and increasing efficiency.

[0112] Based on any of the above embodiments, in Embodiment 5 of the present application, refer to Figure 5 , Figure 5 This is a flowchart of the fifth embodiment of the control method for integrating and scheduling a high-efficiency refrigeration room system in this application. Step S20 includes steps D11 to D13:

[0113] Step D11: Based on the operating data, perform energy efficiency analysis on each subsystem to determine the energy efficiency ratio of each subsystem to the overall system.

[0114] In this embodiment, energy efficiency analysis is to evaluate the energy utilization efficiency of equipment or systems through a mathematical model and identify inefficient operating conditions. The energy efficiency ratio is the ratio of the energy efficiency of each subsystem to the energy efficiency of the entire system.

[0115] As an alternative implementation, based on the operating data of subsystems such as the cooling tower, chilled water pump, and chiller of a central air conditioner, such as the inlet and outlet temperatures of cooling water, the flow rate and power of the water pump, and the compressor COP (Coefficient of Performance) value, first calculate the energy efficiency ratio by subsystem. For example, the energy efficiency of the cooling tower = heat dissipation ÷ power consumption, and the energy efficiency of the chiller = refrigeration capacity ÷ input power. Then, compare the energy efficiency values of each subsystem with the design value or the historical optimal value to obtain the deviation degree. For example, the current energy efficiency of the cooling tower is only 80% of the design value. Finally, integrate the weighted energy efficiency of all subsystems. For example, the weight of the cooling tower is 30% and the weight of the chiller is 50%, and calculate the energy efficiency ratio of the overall system by weighted summation. For example, the energy efficiency ratio of the overall system = energy efficiency of the cooling tower × deviation degree of the cooling tower × weight of the cooling tower + energy efficiency of the chiller × deviation degree of the chiller × weight of the chiller. Finally, calculate the ratio of the energy efficiency ratio of each subsystem to the energy efficiency ratio of the overall system to obtain the energy efficiency ratio of each subsystem to the overall system.

[0116] Step D12: Perform weighted fitting based on the rated energy consumption of the edge devices in each subsystem, and combine the allocation weights to obtain the energy efficiency priority coefficient of each subsystem.

[0117] In this embodiment, the rated energy consumption is the energy consumption calibration value per unit time of the equipment under standard operating conditions. Weighted fitting is to superimpose the rated energy consumption of the edge devices by linear combination or non-linear model according to the weights to generate a comprehensive index. The energy efficiency priority coefficient is a numerical index that quantifies the urgency of optimizing the energy efficiency of the subsystem and is calculated from the ratio of weighted energy consumption to the energy efficiency standard.

[0118] As an alternative implementation, obtain the energy efficiency standard by weighted fitting the rated energy consumption of the edge devices in each subsystem according to the preset energy consumption weights. For example, calculate the energy efficiency standard of each device according to the formula "energy efficiency standard = rated energy consumption × fitting weight", then combine the allocation weights of each subsystem to perform weighted summation on the energy consumption of each type of edge device to obtain the total weighted energy consumption base number, and finally calculate the energy efficiency priority coefficient by "total weighted energy consumption base number ÷ actual energy consumption of the subsystem".

[0119] Step D13: Based on the energy efficiency priority coefficient, the sensor data, and in combination with the comfort coefficient and the energy efficiency ratio, generate the energy efficiency index.

[0120] In this embodiment, the allocated weights will change dynamically according to comfort coefficients, user feedback information, etc., and an adaptive trade-off between energy efficiency and comfort is achieved through dynamic weights.

[0121] As an alternative implementation, first, calculate the current comfort coefficient through sensor data such as temperature, humidity, and wind speed. Then, through the formula comfort compensation value = current comfort - comfort associated with this area. Next, based on the energy efficiency priority coefficient, sensor data, and comfort coefficient compensation value, perform normalization processing on the three types of data, for example, unify them into values between 0 and 1. Then, calculate the comprehensive energy efficiency index according to the dynamic weights, such as the energy efficiency ratio accounting for 50%, the comfort compensation value accounting for 30%, and the priority coefficient accounting for 20%.

[0122] Exemplarily, a central air-conditioning system collects the operating current, valve opening, and energy consumption data of chillers, cooling towers, and fan coils in real-time through edge-side centers deployed in each area. Combining the environmental data of temperature and humidity sensors, carbon dioxide sensors, and infrared passenger flow sensors, the data is aggregated to the cloud. The sliding window algorithm is used to calculate the energy efficiency ratio and regional comfort coefficient of each subsystem. Then, based on the rated energy consumption of the equipment, weighted fitting, and dynamic adjustment factors, the energy efficiency priority coefficient of the subsystem is generated. The global energy efficiency index is generated by fusing the energy efficiency priority coefficient and comfort coefficient through a fuzzy rule engine, and the genetic algorithm is triggered to solve the optimal control parameter combination.

[0123] Since energy efficiency analysis is jointly performed on the sensor data associated with each area, combined with the comfort coefficients of each subsystem and each area, the effect of reducing the energy consumption of the cold source system is achieved without reducing the user's comfort, improving the refrigeration effect and energy-saving effect of the system.

[0124] Based on any of the above embodiments, in the sixth embodiment of the present application, refer to Figure 6 , Figure 6 This is the schematic flowchart of the sixth embodiment of the control method for the integrated scheduling high-efficiency refrigeration machine room system of the present application. Before step S30, steps E11 to E13 are further included:

[0125] Step E11: Based on the sensor data and the comfort coefficient, perform correlation analysis in combination with user feedback information to build a cooling capacity prediction model.

[0126] In this embodiment, the cooling capacity prediction model is a cooling capacity prediction model built based on the correlation analysis of data from each dimension and user feedback information without sacrificing comfort.

[0127] As an alternative implementation, based on sensor data and comfort coefficients, combined with historical user feedback data, first construct a training set through timestamp alignment and data cleaning, then use the Pearson correlation coefficient to screen key features, and then use a long short-term memory neural network model to train a prediction model with time series features and user feedback data labels. Finally, through the detection during training and the fine-tuning of preset parameters, a cooling capacity prediction model is obtained.

[0128] Step E12, the cooling capacity prediction model is trained and verified based on historical cooling capacity demand samples using a multiple linear regression algorithm to generate a cooling capacity demand prediction model.

[0129] In this embodiment, the cooling capacity demand prediction model mainly takes the temperature difference between the supply and return water of the chilled water, the flow rate, and the historical cooling capacity demand as the main features, and at the same time combines external features such as the climate environment (weather, indoor and outdoor air temperature and humidity, wind speed), the population density, and the time period sequence, and selects a multiple linear regression algorithm to train, model, and verify to generate a cooling capacity demand prediction model.

[0130] As an alternative implementation, based on historical cooling capacity demand samples, the system will first clean the data and normalize the feature variables, for example, unify them to 0-1 values, and then divide the training set and the verification set in a ratio of 7:3. Fit the relationship between the features and the cooling capacity demand through a multiple linear regression algorithm (the formula is such as cooling capacity demand = a×temperature difference + b×population density + c×equipment efficiency + intercept term, where a, b, and c are constant terms), use the least squares method to optimize the coefficients and calculate the R² (goodness of fit) score and residual analysis to verify the reliability of the model, and export the cooling capacity demand prediction model after training.

[0131] Step E13, the cooling capacity demand prediction model calculates the cooling capacity prediction data for the current environment based on the operation data.

[0132] In this embodiment, the future cooling capacity demand value output by the model is used to pre-adjust the operation state of the equipment, or it can also be a future 24-hour cooling capacity demand prediction curve output after model calculation.

[0133] As an alternative implementation, real-time collect the operation data of each subsystem, such as the cooling water flow rate, the power of the chilled water pump, and the set value of the terminal temperature. First, perform data cleaning, filter abnormal sensor jumps, fill in missing time periods, and standardize the data, such as unifying the temperature to degrees Celsius and converting the power to kilowatts. Then, input the processed data into a pre-trained cooling capacity demand prediction model, combine the historical load curve and external variables, such as weather forecasts and human flow predictions, and calculate the cooling capacity demand prediction value for the future time period through the internal weights of the model (the formula is such as cooling capacity demand = a×temperature difference + b×population density + c×equipment efficiency + intercept term, where a, b, and c are constant terms).

[0134] Exemplarily, by collecting the operation data of the central air-conditioning system equipment, sensor data, and user feedback information in real time, after data cleaning, by mining the statistical correlation between variables, a multiple linear regression model is constructed based on historical cooling demand samples, the regression coefficients are optimized through cross-validation, and the feature weights are quantified to generate a cooling demand prediction model to calculate the future cooling demand in real time. Combining the prediction results, pre-control instructions for the equipment are dynamically generated, and an encrypted communication protocol is used to ensure the secure transmission of control instructions. The effectiveness of the strategy is verified through a digital twin system, and finally, multi-objective collaborative control of improving the matching degree of cooling supply and demand, reducing energy consumption, and optimizing user comfort is achieved.

[0135] Since a cooling demand prediction model is trained and training samples are collected for its training and optimization, the system can quickly predict the future cooling demand through the cooling demand prediction model, make preparations for equipment response and control strategies in advance, and improve the response speed and scheduling efficiency of the system.

[0136] Based on any of the above embodiments, in the seventh embodiment of the present application, with reference to Figure 7 , Figure 7 is a schematic flowchart of the control method of the seventh embodiment of the integrated scheduling efficient refrigeration machine room system of the present application. Step S30 includes steps F11 to F14:

[0137] Step F11, perform a particle swarm optimization algorithm based on each of the subsystems, and combine the energy efficiency index and the cooling prediction data to obtain a number of global optimal solutions.

[0138] In this embodiment, the particle swarm optimization algorithm is a global optimization algorithm based on swarm intelligence. By simulating the foraging behavior of a bird flock, the particle positions and velocities are iteratively updated to find the multi-objective optimal solution set. The global optimal solution is the solution set that satisfies the constraint conditions and the fitness reaches the preset threshold in multi-objective optimization.

[0139] As an optional implementation manner of the particle swarm optimization algorithm, the particle swarm optimization algorithm is used. In the present invention, the cooling subsystem, the refrigeration subsystem, and the chiller in the above dimensions are used as three particle swarms, and their local optimal solutions and global optimal solutions are calculated respectively. The range of each input variable is constrained, and the standard particle swarm optimization algorithm is used for calculation.

[0140] Step F12, obtain the target global optimal solution by comparing and judging each of the global optimal solutions.

[0141] In this embodiment, the target global optimal solution is obtained by comparing and judging each global optimal solution, and the solution set with the highest fitness is selected as the target global optimal solution.

[0142] As an alternative implementation, the system first collects multiple globally optimal solutions calculated by different optimization algorithms, scores each solution based on a preset weight rule, then compares the scores to select the solution with the highest total score as the target solution, and at the same time verifies whether this solution meets the actual constraint conditions. After passing the verification, it is determined as the target globally optimal solution.

[0143] Step F13, parse the target globally optimal solution to obtain the target local optimal solutions of the corresponding subsystems.

[0144] In this embodiment, the target local optimal solution is a parameter combination that optimizes the fitness function of each subsystem under the condition that the global constraint conditions are satisfied at the entire system level.

[0145] As an alternative implementation, the globally optimal solution of the entire system is disassembled into local optimization problems of each subsystem by the decomposition and coordination method, a distributed optimization algorithm is used to iteratively solve the local optimal solutions of each subsystem, and through Pareto front screening and model predictive control for dynamic adjustment. By judgment, if running according to this local optimal solution can ensure that the local solution meets the energy efficiency index and comfort coefficient, it is defined as the target local optimal solution; if not, readjustment is performed.

[0146] Step F14, generate the control strategies of the corresponding subsystems according to the target local optimal solutions.

[0147] In this embodiment, the control strategies for specifically controlling the edge devices are generated according to the target local optimal solutions, and each edge device is fused and scheduled to execute the corresponding control strategies.

[0148] As an alternative implementation, based on the target local optimal solution, for example, the air-conditioning subsystem needs to maintain 26°C ± 0.5°C and the energy efficiency ratio is increased by 12%, and the frequency of the chilled water pump is adjusted to 45 Hz. First, convert the mathematical optimization parameters, such as the temperature set value and the frequency adjustment step, into control instructions recognizable by the device, and combine the real-time operating status of the subsystem, such as the current temperature deviation, the water pump load rate, and the safety thresholds, such as the voltage fluctuation limit and the compressor start-stop interval, and generate specific control strategies through the rule engine, such as reducing the compressor power by one gear in this area.

[0149] Exemplarily, the particle swarm optimization algorithm is used to perform multi-objective collaborative optimization on each subsystem. Taking the energy efficiency index as the fitness function, multi-dimensional solution spaces are constructed by integrating the cooling capacity prediction data. Through iterative search, a Pareto optimal solution set that meets the maximization of energy efficiency and the balance of cooling capacity supply and demand is generated. The non-dominated solution screening mechanism is combined with business constraints for comparison to determine the global optimal solution. The global solution is mapped to the subsystem level through the decomposition and coordination method. The Lagrange multiplier is used to handle the coupling constraints and analyze the local optimal parameter combinations of the corresponding subsystems. Based on the local optimal structure, the device control strategy is constructed. The optimization parameters are converted into executable instructions through the protocol conversion engine. The digital twin system is synchronously deployed for strategy pre-verification and dynamic compensation mechanism, and finally the closed-loop collaboration of system-level energy efficiency optimization and device-level fine control is achieved.

[0150] Due to the optimization of the value definition and calculation process of each variable, as well as factors such as the verification of the output, the global optimal solution is finally obtained through methods such as polynomial judgment and comparison. The global optimal solutions corresponding to each dimension of the decomposed global optimal solution are obtained, and then the entire calculation process is carried out and formed into a closed loop by fitting and converting each dimension to the parameter level. This improves the matching degree of cooling capacity supply and demand and reduces the energy consumption cost, enhancing the operation efficiency and energy-saving effect of the system.

[0151] Based on any of the above embodiments, in the eighth embodiment of the present application, referring to Figure 8 , Figure 8 is a schematic flowchart of the control method of the eighth embodiment of the integrated scheduling high-efficiency refrigeration machine room system of the present application. Step S40 includes steps G11 to G13:

[0152] Step G11, based on the control strategies corresponding to each subsystem, each edge control strategy corresponding to each region is generated through each edge-end center.

[0153] In this embodiment, the edge control strategy is an edge-end device control rule dynamically generated for a specific region.

[0154] As an optional implementation manner, based on the control strategies of each subsystem, the edge-end center will first receive the superior policy instructions, and at the same time integrate the real-time sensor data of this region and the status of the edge-end devices in the region, and perform policy adaptation through the rule engine or lightweight reinforcement learning model to ensure that the edge-end devices can generate executable edge control strategies under conditions such as not exceeding the rated power.

[0155] Step G12, reversely decompose the edge control strategy to generate the control parameters of the edge-end devices in each region, and send the control parameters to the corresponding edge-end devices.

[0156] In this embodiment, the control strategy is transmitted to the corresponding edge-end device through the Internet of Things connection channel.

[0157] As an alternative implementation, the edge-side central unit will, according to the overall control strategy, first break down the operation requirements into specific device requirements by region, and then automatically convert the parameters that each device needs to adjust into control parameters recognizable by the device through the edge computing nodes, and securely push them to the edge-side devices in the corresponding region through the Internet of Things channel.

[0158] Step G13, the edge-side device operates based on the corresponding control parameters.

[0159] In this embodiment, the control parameters are associated with the device numbers of the corresponding edge-side devices, which is convenient for the edge-side central unit to send the control parameters to the corresponding edge-side devices.

[0160] As an alternative implementation, after receiving the control parameters sent by the edge-side central unit, the edge-side device will first automatically parse the instructions and verify the permissions and formats, then adjust the internal operation mode, and at the same time, continuously monitor whether its own state matches the target parameters. If the deviation is too large, it will immediately send back an alarm message. When operating normally, it will regularly report the execution effect to the edge-side central unit. If a network interruption occurs, it will maintain the basic operation according to the preset "guaranteed strategy".

[0161] Exemplarily, the edge-side central units in each region are driven by the subsystem-level control strategy to generate edge control strategies adapted to the local area. Based on the regional device topology and real-time operation status, the dynamic load balancing algorithm and the constraint satisfaction model are used to reverse-analyze the strategies for the device-level control parameters. The parameters are encapsulated into executable instructions for the device through the protocol conversion engine. Then, after the edge-side device executes the instructions, it will collect feedback data in real time and trigger a compensation mechanism according to the feedback data.

[0162] As an alternative implementation of the closed-loop control architecture for an energy-efficient refrigeration plant system with integrated scheduling, refer to Figure 9 , Figure 9This is the architecture diagram of the closed-loop control of the integrated scheduling high-efficiency refrigerating machine room system in this application. The closed-loop control architecture of the integrated scheduling high-efficiency refrigerating machine room system takes data preprocessing as the core starting point, covering equipment management (such as equipment registration and status monitoring), multi-dimensional configuration (parameter threshold setting), and energy efficiency analysis (historical energy consumption pattern mining). The processed data drives the cooling capacity prediction model (based on time-series load analysis), comfort evaluation model, and energy efficiency evaluation module (energy efficiency index generation) through the model and algorithm iteration verification module (including data synchronization and version control). Subsequently, the particle swarm optimization algorithm is used to fuse multi-dimensional objectives (energy efficiency index, comfort, cooling capacity demand) to generate a dynamic control strategy, which is transmitted to the execution module (protocol conversion through I / O board cards) to link the cooling subsystem (cooling tower fan speed regulation), chiller (compressor frequency regulation), and refrigeration subsystem (water valve opening control). At the same time, the acquisition module uses the sensor network (temperature and humidity, flow, current) to real-time feedback the equipment status and environmental data to the preprocessing layer, forming a closed-loop control loop of "data acquisition - strategy optimization - execution regulation - effect feedback". The overall architecture is clear at all levels, and the modules are coupled through standardized interfaces to achieve full-link intelligent management and control from strategy generation to precise execution of physical equipment.

[0163] As an optional implementation manner of the integrated scheduling high-efficiency refrigerating machine room system module, refer to Figure 10 , Figure 10 This is the flow chart of the integrated scheduling high-efficiency refrigerating machine room system module in this application. The execution process of the integrated scheduling high-efficiency refrigerating machine room system module is that the acquisition module collects data of various dimensions by transforming and installing sensors on various mechanical and electrical equipment in the cold station.

[0164] After the cooling capacity prediction module collects data of various dimensions and conducts correlation analysis with the user feedback data, a cooling capacity prediction model is built on the premise of not sacrificing comfort. The cooling capacity demand prediction model mainly takes the temperature difference between the supply and return water of the chilled water, flow rate, and historical cooling capacity demand as the main features, and at the same time combines external features such as climate environment (weather, indoor and outdoor air temperature and humidity, wind speed), population density, and time series. The multiple linear regression algorithm is selected for training, modeling, and verification to generate the cooling capacity demand prediction model. When the model is applied, data such as the temperature difference between the supply and return water of the chilled water, flow rate, climate environment data for the next 24 hours, and current population density are input, and the cooling capacity demand prediction curve for the next 24 hours is output after model calculation. Every day, the actual cooling capacity curve data and the cooling capacity demand prediction curve data are fused as the model iteration training data for model iteration training.

[0165] The evaluation module needs to build a comfort evaluation model and an energy efficiency evaluation analysis. The comfort evaluation model uses indicators such as temperature, humidity, wind speed, etc., and human comfort feedback as samples to train its relatively general model, and pre-sets this model in the system. By installing indoor temperature, humidity, and wind speed sensors, the corresponding environmental indicators are collected as basic features, and the basic features are converted into derivative features, which are: equivalent temperature, wind chill index, temperature-humidity entropy; and a comfort evaluation model is generated through a random forest algorithm. Basic features: temperature, humidity, wind speed, derivative features: equivalent temperature (Teq = Ta - 0.4(RH - 50)), where, equivalent temperature (Teq, Thermal Equivalent), air temperature (Ta, Air Temperature), relative humidity (RH, Relative Humidity), wind chill index (WCI = 13.12 + 0.6215Ta - 11.37v0.16 + 0.3965Tav0.16), where, wind chill index (WCI, Wind Chill Index), wind speed (v, Wind Speed), temperature-humidity entropy (S = Ta×RH / 100), where, temperature-humidity entropy (S, Summer Heat-Humidity Index). The system real-time takes the environmental indicators collected by each sensor as input and calculates the comfort coefficient of the current environment in real-time. The energy efficiency evaluation analysis includes and respectively conducts energy efficiency analysis on the cooling subsystem, refrigeration subsystem, and chiller. Since intelligent electric meters are installed in each subsystem respectively, through the actual real-time cooling output and real-time power, the energy efficiency ratio of each subsystem and the whole can be obtained in real-time. According to the rated energy consumption of the electromechanical equipment covered in each subsystem, weighted fitting is carried out to assign an energy efficiency priority coefficient to each subsystem. When each subsystem participates in global optimization subsequently, the fitted and weighted indicators are adopted.

[0166] The strategy module is the core module of the system that integrates each subsystem, conducts global optimization analysis, and formulates control strategies. This module needs to fuse the optimal paths in multiple dimensions, such as: cooling subsystem, refrigeration subsystem, chiller, cooling load demand prediction, comfort evaluation, energy efficiency evaluation, etc., as input. The fusion strategy module uses the particle swarm optimization algorithm. In the present invention, the cooling subsystem, refrigeration subsystem, and chiller in the above dimensions are used as three particle swarms, and their local optimal solutions and global optimal solutions are calculated respectively. Range constraints are imposed on each input variable. The standard particle swarm optimization algorithm is used for calculation, and its algorithm is as follows:

[0167]

[0168]

[0169] Among them, is called the inertia factor, and is called the acceleration constant, generally taking [0, 4]. and are random numbers, and the value range is [0, 1]. represents the d-th dimension of the individual extreme value of the i-th variable. represents the d-th dimension of the global optimal solution. represents the updated state value of the system on the d-th dimension / component at the i-th iteration / moment. represents the state value of the system on the d-th dimension / component at the (i - 1)-th iteration / moment. Through the above algorithm, the individual optimal solution and the global optimal solution are calculated in real time for the above five dimensions respectively.

[0170] For these three dimensions, the real-time power data can be extracted in real time from the smart meter devices independently in charge of the three dimensions, that is, the initial values of the X variable in the three dimensions.

[0171] The non-particle coefficient takes the following values:

[0172] and : The initial value is 1.5;

[0173] and : Generated by the rand() function;

[0174] and : Then it is judged whether to update through the fitness function evaluation;

[0175] In the algorithm, w is the inertia factor. In order to stably converge to the global optimal solution, in the selection of the inertia factor, a relatively common linear decrease is used for iteration.

[0176] t

[0177] Among them, is the inertia weight of the i-th particle at the t-th iteration, , are the maximum and minimum values of the inertia weight, and N is the maximum number of iterations set initially (initially set to 2000).

[0178] Through the above algorithm, the power of the three dimensions is calculated respectively, and stored and verified. The verification is calculated through the fitness function, and the fitness function Fitness is as follows:

[0179] Fitness = +

[0180] Among them, is the sum of the energy consumption of the three-dimensional subsystems,

[0181] That is ;

[0182] is the comfort penalty term, that is, the standard deviation of the comfort level from the target value, and is the weight coefficient of the comfort level, and this coefficient is dynamically adjusted according to factors such as season, day and night, sunny and rainy days, etc. represents the cooling power, which is the power consumption for reducing the ambient temperature. represents the refrigeration power, which is the power consumption for deep cooling (below 0°C). represents the total power of the unit, which is the power consumption for the overall operation of the unit;

[0183] This penalty term mainly considers the fluctuations of indoor temperature and humidity, and its conversion algorithm is as follows:

[0184]

[0185] is the equipment performance penalty term, which mainly analyzes the energy efficiency curves of the chiller and the water pump and the load of the current equipment to obtain the value of this term. is the weight coefficient of this term;

[0186] is the parameter safety range penalty term, that is, it specifies the legal range of the parameters of each device. If it is not within the range, the penalty term is strengthened for correction. is the weight coefficient of this term. represents the sum of the squares of the changes in indoor temperature, represents the sum of the squares of the changes in relative humidity.

[0187] Regardless of whether the chiller is of the centrifugal, screw or magnetic levitation type, each device has its high-efficiency operating load range. When the current real-time load of the device is not within the high-efficiency load interval, a device performance penalty value will be generated. After the energy consumption data after each particle swarm iteration passes through the fitness function, it is judged whether it is the individual optimal or global optimal solution. If so, the corresponding variables are updated, that is (individual optimal) or (global optimal).

[0188] There are two situations in which this algorithm can end its iteration process:

[0189] (1)By calculating the global coefficient of performance (COP) in real time, when the difference between adjacent pairs of COP in the last 30 iterations is less than 1e-4, the particle swarm iteration can be terminated;

[0190] (2)When the number of iterations is greater than or equal to the initially set value of N, the particle swarm iteration can be terminated;

[0191] Through optimizing the value definition of each variable, the calculation process, and factors such as output verification, and by means of polynomial judgment and comparison, the global optimal solution is finally obtained, and the global optimal solutions corresponding to each dimension of the decomposed global optimal solution are obtained. Then, through fitting from each dimension to the parameter level, the entire calculation process is connected and a closed loop is formed.

[0192] The global optimal solution is transformed into the execution module. The execution module needs to decompose it into the parameters of each control unit as follows: the control parameters corresponding to the cooling control unit: the frequency of the cooling tower fan, the frequency of the cooling water pump, the degree of opening of the waterway solenoid valve, and the number of cooling towers increased or decreased; the control parameters corresponding to the refrigeration control unit: the frequency of the refrigeration water pump, the degree of opening of the waterway solenoid valve; the control parameters corresponding to the chiller unit: the set value of the chilled water supply temperature, the set value of the chilled water temperature difference, and the number of chillers increased or decreased; the power consumption of the three subsystem dimensions is converted into parameters. Since the relationships of equipment of each model will be different, it is necessary to collect and accumulate historical data to fit their relationships. After the data fitting in the early stage, the global optimal power consumption data of each subsystem can be converted into equipment parameters and sent to the control module.

[0193] The core of the control module is to issue the instructions and parameters required by each control object transmitted by the execution strategy module. At the same time, it ensures the atomic execution of the execution actions of the control objects and can be rolled back in real time. In the previous subsection, the global optimum is converted into the parameters of each control device, and the control instructions are issued in different ways.

[0194] The feedback module includes device and sensor feedback data and user feedback data. The device and sensor feedback data mainly focuses on whether the changes in environmental indicators after the adjustment of device parameters converge to the target value. The user feedback data is mainly used to facilitate the system to obtain negative samples of system iteration in a timely manner during the continuous operation of the system.

[0195] As an alternative implementation method at both the cloud and edge ends, the system operates in a coordinated manner at both the cloud and edge ends. The cloud focuses on the whole-process management and maintenance of device digitization, as well as the relevant configurations required for business applications. In addition, the cloud is responsible for the preprocessing after data sample collection, as well as the verification of the historical cooling capacity prediction model and the optimization of the global optimization algorithm. The edge-side hub is deployed on-site in the computer room, responsible for communicating with various mechanical and electrical equipment and IoT sensors, and issuing control instructions to each edge-side device. In addition, the edge-side hub will regularly synchronize the iterated models and algorithms from the cloud to the local side, and the actual parameters on-site are also calculated on this side through the models and algorithms. The edge-side hub can operate independently of the cloud.

[0196] As an alternative implementation method of the edge-side device structure, refer to Figure 11 , Figure 11 which is the structure diagram of the edge-side device of this application. For edge-side devices with IoT capabilities, they go through a local area network switch and then to the data terminal unit (DTU, Data Terminal Unit), and then the DTU issues parameter instructions to the edge-side device, such as Figure 11 the blue dotted line path in. For edge-side devices without IoT capabilities, they go through the I / O board inside the edge side and then issue signals to the edge-side device, such as Figure 11 the orange dotted line path in. Among them, edge-side devices with IoT capabilities include chillers and intelligent water meters, and edge-side devices without IoT capabilities include temperature sensors, temperature and humidity sensors, wind sensors, pressure difference switches, frequency converters, and relay switches, etc. In the past, the control logic and instruction issuance of edge-side devices in the industry were all through DDC or PLC-related devices to interact with the devices, but in the present invention, the original mode will be replaced by the above two methods to make its control have better scalability.

[0197] Due to the analysis and issuance of the control strategy for the subsystem, the closed-loop precise control of the global strategy - regional instructions - device actions is realized, achieving multi-objective collaborative optimization and improving the overall collaborative efficiency and operation efficiency of the system.

[0198] It should be noted that the above examples are only for understanding this application and do not constitute a limitation to the method for drone inspection based on a smart pole in this application. Based on this technical concept, more forms of simple transformations are within the protection scope of this application.

[0199] This application provides a device for a fusion scheduling high-efficiency refrigeration machine room. The device for a fusion scheduling high-efficiency refrigeration machine room includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein, the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the control method of the fusion scheduling high-efficiency refrigeration machine room system in the first embodiment above.

[0200] Reference will now be made to Figure 12 , which shows a schematic structural diagram of a fusion scheduling high-efficiency refrigeration machine room device suitable for implementing the embodiments of the present application. The fusion scheduling high-efficiency refrigeration machine room device in the embodiments of the present application may include, but is not limited to, mobile terminals such as refrigeration machine rooms, laptop computers, digital broadcast receivers, personal digital assistants (PDAs), tablet computers (PADs), portable multimedia players (PMPs), in-vehicle terminals (such as in-vehicle navigation terminals), etc., and fixed terminals such as digital TVs, desktop computers, etc. Figure 12 The fusion scheduling high-efficiency refrigeration machine room device shown is merely an example and should not impose any limitations on the functions and usage scope of the embodiments of the present application.

[0201] As Figure 12 shown, the fusion scheduling high-efficiency refrigeration machine room device may include a processing device 1001 (such as a central processing unit, a graphics processing unit, etc.), which may perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 1002 or a program loaded from a storage device 1003 into a random access memory (RAM) 1004. In the random access memory 1004, various programs and data required for the operation of the fusion scheduling high-efficiency refrigeration machine room device are also stored. The processing device 1001, the read-only memory 1002, and the random access memory 1004 are connected to each other through a bus 1005. An input / output (I / O) interface 1006 is also connected to the bus. Generally, the following systems may be connected to the I / O interface 1006: an input device 1007 including, for example, a touch screen, a touch pad, a keyboard, a mouse, an image sensor, a microphone, an accelerometer, a gyroscope, etc.; an output device 1008 including, for example, a liquid crystal display (LCD), a speaker, a vibrator, etc.; a storage device 1003 including, for example, a magnetic tape, a hard disk, etc.; and a communication device 1009. The communication device 1009 may allow the fusion scheduling high-efficiency refrigeration machine room device to communicate with other devices wirelessly or wiredly to exchange data. Although the figure shows a fusion scheduling high-efficiency refrigeration machine room device having various systems, it should be understood that it is not required to implement or have all the systems shown. More or fewer systems may be alternatively implemented or had.

[0202] In particular, according to the embodiments disclosed in the present application, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, the embodiments disclosed in the present application include a computer program product, which includes a computer program carried on a computer-readable medium, and the computer program contains program codes for executing the methods shown in the flowcharts. In such an embodiment, the computer program can be downloaded and installed from a network through a communication device, or installed from a storage device 1003, or installed from a read-only memory 1002. When the computer program is executed by a processing device 1001, the above-mentioned functions defined in the methods of the embodiments disclosed in the present application are executed.

[0203] The integrated scheduling high-efficiency refrigeration machine room equipment provided by the present application adopts the control method of the integrated scheduling high-efficiency refrigeration machine room system in the above-mentioned embodiments, and can solve the technical problem of poor refrigeration effect of the central air conditioner when scheduling the cold source machine room. Compared with the prior art, the beneficial effects of the integrated scheduling high-efficiency refrigeration machine room equipment provided by the present application are the same as those of the control method of the integrated scheduling high-efficiency refrigeration machine room system provided by the above-mentioned embodiments, and other technical features in the integrated scheduling high-efficiency refrigeration machine room equipment are the same as the features disclosed in the method of the previous embodiment, and will not be elaborated here.

[0204] It should be understood that each part disclosed in the present application can be implemented by hardware, software, firmware or a combination thereof. In the description of the above embodiments, specific features, structures, materials or characteristics can be combined in a suitable manner in any one or more embodiments or examples.

[0205] As mentioned above, the above are only specific embodiments of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art within the technical scope disclosed in the present application can easily think of changes or substitutions, which should all be covered by the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

[0206] The present application provides a computer-readable storage medium, having computer-readable program instructions (i.e., computer programs) stored thereon, and the computer-readable program instructions are used to execute the control method of the integrated scheduling high-efficiency refrigeration machine room system in the above-mentioned embodiments.

[0207] The computer-readable storage medium provided by the present application can be, for example, a USB flash drive, but is not limited to electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices, or components, or any combination of the above. More specific examples of the computer-readable storage medium may include, but are not limited to: electrical connections with one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM), or flash memory, optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the above. In this embodiment, the computer-readable storage medium can be any tangible medium that contains or stores a program that can be used by or in conjunction with an instruction execution system, device, or component. The program code contained on the computer-readable storage medium can be transmitted using any appropriate medium, including but not limited to: wires, optical cables, radio frequency (RF), etc., or any suitable combination of the above.

[0208] The above computer-readable storage medium can be included in the integrated scheduling and highly efficient refrigeration machine room equipment; it can also exist independently and not be assembled into the integrated scheduling and highly efficient refrigeration machine room equipment.

[0209] The above computer-readable storage medium carries one or more programs. When the above one or more programs are executed by the integrated scheduling and highly efficient refrigeration machine room equipment, the integrated scheduling and highly efficient refrigeration machine room equipment is caused to: obtain the operation data of each edge device and the sensor data in each region; calculate the energy efficiency indicators of each subsystem based on the operation data, the allocation weights of the subsystems corresponding to each device type, the comfort coefficients associated with each region, and the sensor data in each region; formulate control strategies for each subsystem according to the energy efficiency indicators in combination with the cooling capacity prediction data associated with each region; decompose the control strategies in reverse to obtain the control parameters of the edge devices in each region, and send the control parameters to the edge devices in the corresponding region.

[0210] Computer program code for performing the operations of this application can be written in one or more programming languages or combinations thereof. The above-mentioned programming languages include object-oriented programming languages such as Java, Smalltalk, C++, and also include conventional procedural programming languages such as the "C" language or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, executed as a stand-alone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer can be connected to the user's computer through any kind of network, including a local area network (LAN) or a wide area network (WAN), or it can be connected to an external computer (for example, by connecting through the Internet service provider via the Internet).

[0211] The flowcharts and block diagrams in the accompanying drawings illustrate the possible architectures, functions, and operations of systems, methods, and computer program products according to various embodiments of this application. In this regard, each block in the flowchart or block diagram can represent a module, a program segment, or a part of the code that contains one or more executable instructions for implementing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the blocks may occur in a different order than that marked in the accompanying drawings. For example, two consecutive blocks shown may actually be executed substantially in parallel, and they may sometimes be executed in the reverse order, depending on the functions involved. It should also be noted that each block in the block diagram and / or flowchart, and the combination of blocks in the block diagram and / or flowchart, can be implemented by a dedicated hardware-based system for performing the specified functions or operations, or can be implemented by a combination of dedicated hardware and computer instructions.

[0212] The modules described in the embodiments of this application can be implemented in software or in hardware. Among them, the name of the module does not constitute a limitation to the unit itself in some cases.

[0213] The readable storage medium provided by this application is a computer-readable storage medium that stores computer-readable program instructions (i.e., computer programs) for performing the control method of the above-mentioned integrated scheduling efficient refrigeration machine room system, and can solve the technical problem of poor refrigeration effect of the central air conditioner when scheduling the cold source machine room. Compared with the prior art, the beneficial effects of the computer-readable storage medium provided by this application are the same as those of the control method of the integrated scheduling efficient refrigeration machine room system provided in the above embodiments, and will not be elaborated here.

[0214] The above are only some embodiments of the present application, and thus do not limit the patent scope of the present application. Any equivalent structural transformation made under the technical concept of the present application by using the content of the specification and drawings of the present application, or any direct / indirect application in other related technical fields, is included in the patent protection scope of the present application.

Claims

1. A control method for integrating scheduling of high-efficiency refrigeration room system, characterized in that: The method comprises: Obtain the operating data of each edge device and the sensor data in each area; Based on the sensor data, extract basic features corresponding to the sensor data, and convert the basic features into derived features; Based on the derived features, a comfort evaluation model is generated by modeling using a random forest algorithm; The comfort evaluation model uses the basic features and user feedback information as training samples, and optimizes the comfort evaluation model through the training samples; The comfort evaluation model calculates the comfort coefficient of the current environment according to the sensor data; Based on the operating data, an energy efficiency analysis is performed on each subsystem to determine the energy efficiency ratio of each subsystem to the overall system; Performing weighted fitting according to the rated energy consumption of the edge devices in each of the subsystems, and combining the allocated weights, to obtain the energy efficiency priority coefficient of each of the subsystems; Generate an energy efficiency index based on the energy efficiency priority coefficient, the sensor data, the comfort coefficient and the energy efficiency ratio; According to the energy efficiency index and in combination with the cooling capacity prediction data associated with each of the areas, a control strategy for each of the subsystems is formulated; Reverse decomposition of the control strategy to obtain control parameters of the edge devices in each of the areas, and send the control parameters to the edge devices in the corresponding areas, wherein the reverse decomposition is to decompose the global control strategy into device-level executable instructions through mathematical inversion or a rule engine.

2. The control method of the integrated scheduling high-efficiency refrigeration room system according to claim 1 is characterized in that: Before the step of obtaining the operating data of each edge device and the sensor data in each area, the step further includes: The edge terminal hub receives the start instruction and parses the corresponding business information according to the start instruction; Based on the service information, configure a logical mapping relationship between the edge device corresponding to the start instruction and the service; Based on the basic parameters of the edge device, the basic parameters are adjusted in combination with the logical mapping relationship to generate policy parameters; The policy parameters are configured on the corresponding edge devices, and the edge devices are controlled to operate according to the corresponding policy parameters.

3. The control method of the integrated scheduling high-efficiency refrigeration room system according to claim 1 is characterized in that: The step of obtaining the operating data of each edge device and the sensor data in each area includes: Acquire the original operation data of each edge device and the original sensor data transmitted by the sensors in each area; Data preprocessing is performed based on the operation raw data and the sensor raw data to generate the operation data and the sensor data, and the operation data and the sensor data are stored in a data pool.

4. The control method of the integrated scheduling high-efficiency refrigeration room system according to claim 1 is characterized in that: Before the step of formulating the control strategy of each of the subsystems according to the energy efficiency index and the cooling capacity prediction data associated with each of the regions, the step further includes: Based on the sensor data and the comfort coefficient, a correlation analysis is performed in combination with user feedback information to build a cooling capacity prediction model; The cooling demand prediction model is based on historical cooling demand samples, and a multivariate linear regression algorithm is trained and verified to generate a cooling demand prediction model; The cooling demand prediction model calculates the cooling demand prediction data of the current environment based on the operating data.

5. The control method of the integrated scheduling high-efficiency refrigeration room system according to claim 4 is characterized in that: The step of formulating the control strategy of each subsystem according to the energy efficiency index and the cooling capacity prediction data associated with each area includes: Based on each of the subsystems, a particle swarm optimization algorithm is performed, and several global optimal solutions are obtained by combining the energy efficiency index and the cooling capacity prediction data; By comparing and judging the global optimal solutions, a target global optimal solution is obtained; Analyze the target global optimal solution to obtain the corresponding target local optimal solution of each of the subsystems; The control strategy of each of the subsystems is generated according to the target local optimal solution.

6. The control method of the integrated scheduling high-efficiency refrigeration room system according to claim 1 is characterized in that: The step of reversely decomposing the control strategy to obtain control parameters of the edge devices in each of the areas, and sending the control parameters to the edge devices in the corresponding areas includes: Based on the control strategies corresponding to the subsystems, edge control strategies corresponding to the regions are generated through the edge-end hubs; Reversely decompose the edge control strategy, generate the control parameters of the edge devices in each of the areas, and send the control parameters to the corresponding edge devices; The edge device operates based on the corresponding control parameters.

7. A fusion-scheduling high-efficiency refrigeration room equipment, characterized in that: The fusion-scheduling high-efficiency refrigeration room equipment includes: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the computer program is configured to implement the steps of the control method of the fusion-scheduling high-efficiency refrigeration room system as described in any one of claims 1 to 6.

8. A storage medium, characterized in that: The storage medium is a computer-readable storage medium, and a computer program is stored on the storage medium. When the computer program is executed by a processor, the steps of the control method of the fusion scheduling high-efficiency refrigeration room system according to any one of claims 1 to 6 are implemented.

Images