Multi-matrix equipment temperature cloud control method

Abstract

The invention discloses a multi-matrix equipment temperature cloud control method, relates to the technical field of equipment temperature control, and solves the technical problem that in the prior art, a temperature control strategy is wrong due to the fact that no timely countermeasure is made to change of an external environment. According to the method, the real-time temperature of the multi-matrix equipment is collected, the operation data of the equipment is extracted, the standard temperature position is set, and the temperature sensor is calibrated to obtain the calibration temperature; calculating and obtaining heat influence data of the connection equipment on the test equipment according to the equipment layout, and calculating and obtaining a unit influence rate to generate an equipment prediction temperature rise map; a prediction target temperature is selected and compared with a standard temperature to obtain an adjustment parameter, the adjustment parameter is input into a temperature control model constructed by the neural network model to obtain a temperature control action strategy, and the intelligent temperature control actuator performs temperature adjustment according to the temperature control action strategy; and detecting temperature state data during temperature regulation and control in real time, and performing secondary optimization to obtain optimized temperature data.

Description

Technical Field

[0001] The invention belongs to the field of device temperature control, and in particular is a multi-matrix device temperature cloud control method. Background Art

[0002] In industrial production (such as electronic chip manufacturing, chemical reaction processes), large data center computer rooms, smart warehousing cold chains and other scenarios, a large number of matrix arrangement devices are distributed, which have extremely high requirements for the accuracy and stability of temperature control. The matrix arrangement equipment will generate a lot of heat during operation. If the heat cannot be effectively dissipated, it will seriously affect the stability and service life of the server.

[0003] The invention patent with application number CN202110778803X discloses a temperature control method and device, a modular data center and a storage medium. The invention monitors the power consumption and temperature status data of the temperature controlled object, and inputs the temperature status data into the temperature control strategy model. In the temperature control measurement model, it includes a reward function, a strategy model, and an environmental model. The environmental model is updated and trained based on the temperature control action. Based on the temperature state at each moment, the temperature control action at the next moment is trained to perform real-time temperature control. In the multi-matrix device layer, the devices are close to each other, so the heat generated during operation will be transferred to each other. When performing temperature control, changes in the external environment and the heat transferred by other devices will affect the temperature control strategy of the device. When this method is training the temperature control action, if the external environment changes suddenly, the subsequent temperature control strategy obtained will not reach the required control temperature, resulting in the problem of temperature control strategy error.

[0004] The present invention provides a multi-matrix equipment temperature cloud control method to solve the above technical problems. Summary of the invention

[0005] The present invention aims to solve at least one of the technical problems existing in the prior art; to this end, the present invention proposes a multi-matrix equipment temperature cloud control method, which is used to solve the technical problem in the prior art that no timely countermeasures are taken to changes in the external environment, resulting in errors in the temperature control strategy.

[0006] To achieve the above object, a first aspect of the present invention provides a multi-matrix device temperature cloud control method, comprising:

[0007] Step 1: Periodically collect the real-time temperature of each multi-matrix device and extract the device operation data of the multi-matrix device from the management database;

[0008] Step 2: calibrate the real-time temperature to obtain the calibration temperature;

[0009] Step 3: Obtain a predicted temperature rise diagram of the equipment based on the calibration temperature and the equipment operation data; obtain adjustment parameters based on the comparison result between the predicted temperature rise diagram of the equipment and the standard temperature in the management database;

[0010] Step 4: Send the adjustment parameters to the temperature controller for temperature control, and continuously send back temperature control status data;

[0011] Step 5: Perform secondary optimization based on the temperature state data and the adjustment parameters to obtain the optimized temperature data, and input the optimized temperature data into the temperature controller for secondary regulation.

[0012] Preferably, calibrating the real-time temperature to obtain the calibration temperature comprises:

[0013] Select a standard temperature position, and place a temperature sensor at the standard temperature position to measure the measured temperature of the standard temperature position, calculate the measured temperature and the standard temperature corresponding to the standard temperature position to obtain a calibration temperature difference; add the real-time temperature and the calibration temperature difference to obtain the calibration temperature.

[0014] Preferably, the predicting of the equipment temperature rise diagram based on the calibration temperature and the equipment operation data includes:

[0015] Extract calibration temperature and equipment operation data; wherein the equipment operation data includes equipment layout, load size, operation time and external environment temperature of multi-matrix equipment;

[0016] Obtain the thermal impact data of multi-matrix equipment according to the equipment layout; define the unit time and the expected temperature that the equipment is expected to reach within the unit time; mark the temperature reached by the test equipment within the unit time as the test temperature, and mark the difference between the expected temperature and the test temperature as the error temperature;

[0017] The unit impact rate WTS is calculated by the formula WTS=(CST-JZT)×WBT×α / (FZR×YXT); wherein CST represents the test temperature, JZT represents the calibration temperature, WBT represents the external ambient temperature, FZR represents the load size, YXT represents the operating time, and α represents the environmental impact coefficient, which is set based on the difference between the external ambient temperature and the test temperature;

[0018] With time as the horizontal axis, the temperature of the equipment as the vertical axis, and the unit impact rate as the slope, a predicted temperature rise graph of the equipment is generated.

[0019] It should be noted that the equipment prediction temperature rise diagram is actually nonlinear. As the load changes, its unit impact rate also changes accordingly. In the equipment prediction temperature rise diagram, the unit impact rate is dynamically adjusted according to the real-time changes in the load, so that the temperature predicted by the equipment preset temperature rise diagram is more accurate.

[0020] Preferably, the step of obtaining the thermal impact data of the multi-matrix equipment according to the equipment layout includes:

[0021] Extracting device layout; wherein the device layout includes the length of cables connecting the multi-matrix devices and the distribution position of each multi-matrix device;

[0022] Marking other multi-matrix devices connected to the multi-matrix device by data cables as connection devices, and marking the multi-matrix device to be subjected to temperature test as test device; measuring the connection device temperature of the connection device by a calibrated temperature sensor, and subtracting the connection device temperature from the calibration temperature to obtain a calibrated connection temperature;

[0023] Determine whether the calibration temperature is greater than the calibration connection temperature; if yes, mark the thermal impact state of the connection device on the test device as a negative impact state; if no, mark the thermal impact state of the connection device on the test device as a positive impact state;

[0024] Convert the distribution position of the multi-matrix device to obtain the distribution distance between the test device and the connected device; by the formula RYX = (JLT-JZT) × ln (FL) / ln (DL) 3 The layout temperature influence RYX is calculated; where JLT represents the calibration connection temperature, JZT represents the calibration temperature, FL represents the distribution distance, and DL represents the cable length;

[0025] The layout temperature impact and thermal impact status are integrated to obtain the thermal impact data.

[0026] Preferably, the environmental impact coefficient is set based on the difference between the external environment temperature and the test temperature, including:

[0027] Extracting several groups of historical test data of the test equipment under known but different external ambient temperatures from the historical database;

[0028] Randomly select a test running time, and mark the test temperature corresponding to the test running time in the historical test data as the historical test temperature; calculate the ratio of the difference between the historical test temperature and the calibration temperature and the difference between the external ambient temperature and the calibration temperature to obtain several groups of test influence coefficients; set the error term, and select the average of several groups of test influence coefficients as the verification coefficient; take the value obtained by subtracting the product of the difference between the external ambient temperature and the calibration temperature and the verification coefficient from the difference between the historical test temperature and the calibration temperature as the error value;

[0029] Determine whether the error value corresponding to the historical test data meets the error term; if yes, retain the corresponding historical test data; if no, remove the corresponding historical test data;

[0030] The mode of the test impact coefficients corresponding to the remaining historical test data is taken as the environmental impact coefficient.

[0031] It should be noted that the error value is set based on the degree of normal instantaneous changes in the external environment, such as instantaneous changes in wind speed or temperature.

[0032] Preferably, the step of obtaining the adjustment parameter based on the comparison result between the equipment predicted temperature rise diagram and the standard temperature in the management database includes:

[0033] Select the target time; obtain the predicted target temperature corresponding to the target time in the equipment predicted temperature rise diagram;

[0034] Determine whether the predicted target temperature is greater than the standard temperature; if yes, mark the adjustment state as a cooling state; if no, mark the adjustment state as a warming state;

[0035] The difference between the predicted target temperature and the standard temperature is marked as the adjustment temperature; the adjustment temperature and the adjustment state are integrated to obtain the adjustment parameter.

[0036] Preferably, sending the adjustment parameters to the temperature controller for temperature control includes:

[0037] Extract adjustment parameters;

[0038] Determine whether the adjustment state is a cooling state; if yes, start the refrigeration element in the intelligent temperature control actuator; if not, start the heating element in the intelligent temperature control actuator; input the adjustment temperature into the temperature control model to obtain the temperature control action strategy, and input the temperature control action strategy into the intelligent temperature control actuator for temperature regulation; wherein, the temperature control model is constructed based on the neural network model.

[0039] Preferably, the temperature control model is constructed based on a neural network model, including:

[0040] Several groups of historical adjustment parameters and corresponding historical temperature control action strategies are extracted from the historical database; the adjustment parameters and the corresponding historical temperature control action strategies are used as training data and test data, the neural network model is trained by the training data, the trained neural network model is tested by the test data, and the parameters of the neural network model are adjusted according to the test results to obtain a temperature control model with the input data as the adjustment parameters and the output data as the temperature control action strategy.

[0041] Preferably, the secondary optimization based on the temperature state data and the adjustment parameters to obtain the optimized temperature data includes:

[0042] A1: Extract temperature status data and adjustment parameters; wherein the temperature status data includes the control temperature of the device after temperature control and the operation data of the control device at the corresponding time;

[0043] A2: Mark the time corresponding to the controlled temperature as the controlled time, and match the predicted temperature corresponding to the controlled time in the device predicted temperature graph;

[0044] A3: Determine whether the controlled temperature is the same as the predicted temperature; if yes, mark the corresponding temperature control state as normal; if no, mark the corresponding temperature control state as a state requiring optimization and jump to A4;

[0045] A4: Determine whether the controlled temperature is lower than the predicted temperature; if yes, mark the optimized state as a state requiring heating; if no, mark the optimized state as a state requiring cooling;

[0046] A5: Based on the temperature control state and the optimization state, the intelligent temperature control actuator is optimized twice to obtain the optimized temperature data.

[0047] Preferably, the second optimization of the intelligent temperature control actuator based on the temperature control state and the optimization state to obtain the optimized temperature data includes:

[0048] For adjustment parameters whose temperature control state is a state that needs to be optimized, the environmental impact coefficient is updated according to the corresponding optimization state and temperature state parameters to obtain the updated impact coefficient; the updated temperature control action strategy and updated adjustment parameters are re-obtained through the updated influence coefficient, and the updated temperature control action strategy and updated adjustment parameters are integrated to obtain the optimized temperature data.

[0049] Compared with the prior art, the present invention has the following beneficial effects:

[0050] 1. The present invention collects the real-time temperature of multi-matrix devices periodically and in real time, and extracts the device operation data of the multi-matrix devices from the management database; by setting the standard temperature position, the temperature sensor is calibrated to obtain the calibration temperature, and the measurement error caused by sensor aging or the problem of the sensor itself is eliminated to ensure the accuracy of the temperature data; the thermal impact data of the connected device on the test device is obtained according to the device layout calculation, and the expected temperature obtained by division is given, and the unit impact rate is calculated, and the device prediction temperature rise diagram is generated based on the unit impact rate, which is helpful to understand the thermal interaction between different devices and eliminate the influence of other devices on the test equipment; the predicted target temperature is selected from the device prediction temperature rise diagram, and compared with the standard temperature to obtain the adjustment parameters that need to be adjusted, and the adjustment parameters are input into the temperature control model obtained by constructing the neural network model to obtain the temperature control action strategy, and the intelligent temperature control actuator performs temperature adjustment according to the temperature control action strategy, which helps to reduce temperature fluctuations and errors and improve the working efficiency and accuracy of the equipment; the temperature state data when adjusting the temperature is detected in real time, and secondary optimization is performed to obtain optimized temperature data, and the intelligent temperature control actuator is optimized and adjusted by optimizing the temperature data to ensure the completion of temperature control.

[0051] 2. The present invention sets the environmental impact coefficient according to the difference between the external ambient temperature and the test temperature, and sets the error term to eliminate the environmental impact caused by normal environmental changes. In combination with the layout between the test equipment and the connection equipment, the layout temperature impact between the connection equipment and the test equipment is calculated. The thermal impact between the test equipment and the connection equipment under different equipment layouts can be quantitatively evaluated, and the thermal impact analysis between the equipment is more accurate. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0053] Figure 1 The following is a flowchart of an embodiment of the present invention.

[0054] Figure 2 This is a diagram of predicted temperature rise of the device obtained in one embodiment of the present invention.

[0055] Figure 3 The figure is a complete flow chart of secondary optimization in one embodiment of the present invention. DETAILED DESCRIPTION

[0056] The technical solution of the present invention will be clearly and completely described below in conjunction with the embodiments. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0057] See also Figure 1-Figure 3 The first aspect of the present invention provides a multi-matrix device temperature cloud control method, comprising:

[0058] Step 1: Periodically collect the real-time temperature of each multi-matrix device and extract the device operation data of the multi-matrix device from the management database.

[0059] Exemplarily, high-precision temperature sensors are respectively set in several multi-matrix devices in a data center, and the real-time temperatures of all the multi-matrix devices are collected every several seconds. The device layout of each multi-matrix device in the data center, the load size in each multi-matrix device, the continuous operating time, and the external ambient temperature collected by the high-precision temperature sensors set in the data center are extracted from the management database of the data center.

[0060] Step 2: Calibrate the real-time temperature to obtain the calibration temperature.

[0061] Exemplarily, a standard temperature position is selected inside the data center. In this embodiment, a standard temperature position with a temperature of 25 degrees is selected in the data center. The temperature sensor is placed at the standard temperature position to measure the temperature of the standard temperature position. If the measured temperature is the same as the standard temperature, it means that the measurement result of the temperature sensor is accurate during measurement, and no calibration is required. If the measured temperature is different from the standard temperature, it means that there is a deviation in the temperature sensor during measurement. The difference between the measured temperature and the standard temperature is used as the calibration temperature difference of the corresponding temperature sensor. The real-time temperature measured by the temperature sensor in the multi-matrix device is added to the calibration temperature difference to obtain the actual measured calibration temperature. The result of the sensor's temperature detection is eliminated, and the measurement error caused by sensor aging or problems with the sensor itself is eliminated, thereby ensuring the accuracy of the temperature data.

[0062] It should be noted that the calibration temperature position is generally selected at the absolute temperature, that is, 0 degrees or 100 degrees, but absolute temperature generally does not appear in a data center, so a position corresponding to a known temperature is selected as the calibration temperature position.

[0063] Step 3: Obtain a predicted equipment temperature rise diagram based on the calibration temperature and equipment operation data prediction; obtain adjustment parameters based on the comparison result between the predicted equipment temperature rise diagram and the standard temperature in the management database.

[0064] Exemplarily, in this embodiment, the temperature of multi-matrix device A is measured, other multi-matrix devices connected to multi-matrix device A by data cables are marked as connection devices, multi-matrix device A is marked as test device, the temperature measured by the temperature sensor of the connection device is marked as the connection device temperature, and calibration is performed in combination with the calibration temperature to obtain the calibrated connection temperature; it is determined whether the calibration temperature is greater than the calibrated connection temperature. In this embodiment, multi-matrix device B connected to multi-matrix device A is selected as the connection device. The calibrated connection temperature of the connection device is greater than the test device, indicating that the temperature of the connection device is higher than the temperature of the test device. When transmitting heat, the connection device will affect the test device and transmit heat to the test device. Therefore, the thermal influence state of the connection device on the test device is marked as a positive influence state, which helps to understand the thermal interaction between different devices, and can also provide a basis for optimizing airflow management and cooling strategies in the data center, and analyze the influence of other connection devices in the data center on the temperature of the test device.

[0065] In some other preferred embodiments, the temperature of the connecting device is lower than that of the testing device. In this case, when transferring heat, the heat of the testing device is higher than that of the connecting device. The testing device will transfer heat to the connecting device, so the thermal impact state of the connecting device on the testing device is marked as a negative impact state. In actual processes, multiple connecting devices are connected to the testing device, and the temperatures of the connecting devices are different. Therefore, there may be several connecting devices in a positive impact state and several connecting devices in a negative impact state connected to the testing device, and different connecting devices have different impacts on the testing device.

[0066] According to the distribution position of the multi-matrix equipment, the distribution distance between the test equipment and the connection equipment can be obtained. Due to the closed space, the multi-matrix equipment will affect each other; through the formula RYX = (JLT-JZT) × ln (FL) / ln (DL) 3 The layout temperature impact RYX is calculated. The closer the distribution distance, the greater the impact between the test equipment and the connecting equipment. The longer the cable between the test equipment and the connecting equipment, the more heat is dissipated during transmission between the test equipment and the connecting equipment, and the smaller the impact. The thermal impact data of the connecting equipment on the test equipment is obtained by integrating different temperature impacts with the thermal impact status of the connecting equipment. This can quantitatively evaluate the thermal impact between the test equipment and the connecting equipment under different equipment layouts, and make the thermal impact analysis between the equipment more accurate.

[0067] In this embodiment, the unit time is defined as 5 minutes, the expected temperature of the test equipment within the unit time is 30 degrees, the temperature obtained by the test equipment after 5 minutes is marked as the test temperature, and the difference between the expected temperature and the test temperature is marked as the error temperature; the setting of the expected temperature needs to be set according to the historical data of the test equipment.

[0068] Extract historical test data of several groups of test equipment from the historical database, in which other data are the same but the external environment temperature is different and known; select the test running time in the historical data, mark the test temperature corresponding to the test running time in the historical test data as the historical test temperature, obtain different external environment temperatures in the historical test data, calculate the difference between the historical test temperature and the calibration temperature to obtain the historical test difference, mark the difference between the external environment temperature and the calibration temperature as the external difference, compare the historical test difference with the difference of the external environment to obtain the corresponding test influence coefficient, and the test influence coefficient can reflect the influence of the external environment temperature on the internal equipment; in this embodiment, an error term is set, and the error term is essentially a range. In actual changes, the external environment is not constant and may be due to The sudden change in temperature leads to errors in the analysis of the external environment; the average of several groups of test influence coefficients is selected as the verification coefficient, and the error value is obtained by subtracting the external difference from the historical test difference and multiplying the result by the verification coefficient; the error value is used to distinguish the coefficient changes caused by normal changes in the external environment when analyzing the historical data; it is judged whether the error value corresponding to the historical test data meets the error term. In this embodiment, the error value obtained meets the error term and belongs to normal environmental changes. The mode of the test influence coefficients corresponding to the remaining historical test data is used as the environmental influence coefficient, which can accurately quantify the degree of influence of the external environmental temperature on the internal equipment test temperature, exclude the coefficient changes caused by normal changes in the external environment, eliminate or reduce the influence of the external environmental temperature on the test results, thereby improving the accuracy of data analysis.

[0069] The unit impact rate WTS of the test equipment is calculated by the formula WTS=(CST-JZT)×WBT×α / (FZR×YXT); wherein CST represents the test temperature, JZT represents the calibration temperature, WBT represents the external ambient temperature, FZR represents the load size, YXT represents the operating time, and α represents the environmental impact coefficient; with time as the horizontal axis, the temperature of the equipment as the vertical axis, and the unit impact rate as the slope, a predicted temperature rise diagram of the equipment is generated.

[0070] In this embodiment, after selecting the target time as 10 minutes after the current time, the predicted target temperature corresponding to the target time is matched in the equipment prediction temperature rise diagram; the standard temperature that the test equipment should reach is extracted from the management database, and it is determined whether the predicted target temperature is greater than the standard temperature. In this embodiment, the predicted target temperature is 30 degrees, while the standard temperature is 32 degrees. The predicted target temperature is less than the standard temperature, indicating that without the action of the temperature controller, the temperature at the target time is less than the standard temperature to be reached. The adjustment state is marked as the warming state, and the difference between the predicted target temperature and the standard temperature is the temperature value that needs to be adjusted. The difference between the predicted target temperature and the standard temperature is marked as the adjustment temperature. The adjustment temperature and the adjustment state are integrated to obtain the adjustment parameter, which can gradually improve the accuracy of temperature control. This helps to reduce temperature fluctuations and errors and improve the working efficiency and accuracy of the equipment.

[0071] In some other preferred embodiments, if the predicted target temperature is greater than the standard temperature, it means that in the absence of a temperature controller, due to the influence of the connecting device on the test device, the temperature of the test device is higher than the standard temperature that needs to be reached, and the adjustment state of the test device is marked as a cooling state.

[0072] Step 4: Send the adjustment parameters to the temperature controller for temperature control, and continuously send back temperature control status data.

[0073] Exemplarily, the adjustment parameters of the test equipment are sent to the temperature controller, and the temperature control element in the intelligent temperature control actuator is started according to the adjustment state. If it is in a cooling state, the refrigeration element is started, and if it is in a warming state, the heating element is started; several groups of historical adjustment parameters and corresponding historical temperature control action strategies are extracted from the historical database; the adjustment parameters and the corresponding historical temperature control action strategies are used as training data and test data, and the neural network model is trained by the training data, and the trained neural network model is tested by the test data. According to the test results, the parameters of the neural network model are adjusted to obtain a temperature control model whose input data is the adjustment parameters and whose output data is the temperature control action strategy; the adjustment parameters of the test equipment are input into the temperature control model as input data to obtain the power size required to be input and the start-up state of the expected temperature control time, and the temperature control action strategy is obtained by integration, and the intelligent temperature control actuator regulates the temperature of the test equipment according to the temperature control action strategy.

[0074] During the regulation process, the equipment operation data of the test equipment and the real-time temperature during regulation are collected in real time; the predicted temperature corresponding to the regulation time after the regulation is completed is matched in the equipment predicted temperature diagram; it is determined whether the regulated temperature is the same as the predicted temperature. In this embodiment, it is determined that the regulated temperature is different from the predicted temperature, which means that the adjustment effect cannot meet the required requirements and the adjustment parameters need to be optimized. The temperature control state of the intelligent temperature control actuator is marked as a state to be optimized, and it is further determined whether the regulated temperature is less than the predicted temperature. In this embodiment, the regulated temperature is 33 degrees and the predicted temperature is 35 degrees. The regulated temperature is less than the predicted temperature, which means that further temperature increase is required and the optimized state is marked as a state to be heated. The environmental influence coefficient is recalculated through the optimization state and the temperature state parameters to obtain an updated influence coefficient, and the temperature control model is re-entered to obtain the updated adjustment parameters and the updated temperature control action strategy. The intelligent temperature control actuator is controlled through the updated temperature control action strategy, and secondary optimization is performed to adjust the temperature of the control test equipment to eliminate the temperature control error caused by changes in the external environment.

[0075] It should be noted that during the temperature control process, the external environment may change temporarily, causing the intelligent temperature control actuator to use data based on the data before the change when adjusting the temperature. Under the influence of the changed external environment temperature, the final temperature is different from the predicted temperature, so secondary optimization is performed to ensure accurate temperature control.

[0076] Part of the data in the above formula is calculated by removing the dimension and taking its numerical value. The formula is a formula closest to the actual situation obtained by software simulation of a large amount of collected data; the preset parameters and preset thresholds in the formula are set by technical personnel in this field according to actual conditions or obtained through simulation of a large amount of data.

[0077] Working principle of the present invention:

[0078] The present invention periodically collects the real-time temperature of multi-matrix equipment in real time, extracts the equipment operation data of the multi-matrix equipment from a management database, and calibrates the temperature sensor based on the set standard temperature position; sets the environmental impact coefficient, and combines the layout between the test equipment and the connection equipment to calculate the layout temperature impact between the connection equipment and the test equipment, calculates the unit impact rate and generates the equipment prediction temperature rise diagram; selects the predicted target temperature in the equipment prediction temperature rise diagram, and compares it with the standard temperature to obtain the adjustment parameter, inputs the adjustment parameter into the temperature control model constructed by the neural network model to obtain the temperature control action strategy, and the intelligent temperature control actuator adjusts the temperature according to the temperature control action strategy; detects the temperature state data when adjusting the temperature in real time, and performs secondary optimization to obtain the optimized temperature data, and optimizes and adjusts the intelligent temperature control actuator through the optimized temperature data to complete the temperature control.

[0079] The above embodiments are only used to illustrate the technical method of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical method of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical method of the present invention.

Claims

1. A multi-matrix device temperature cloud control method, characterized in that: include: Step 1: Periodically collect the real-time temperature of each multi-matrix device and extract the device operation data of the multi-matrix device from the management database; Step 2: calibrate the real-time temperature to obtain the calibration temperature; Step 3: Obtain a predicted temperature rise diagram of the equipment based on the calibration temperature and the equipment operation data; obtain adjustment parameters based on the comparison result between the predicted temperature rise diagram of the equipment and the standard temperature in the management database; Step 4: Send the adjustment parameters to the temperature controller for temperature control, and continuously send back temperature control status data; Step 5: Perform secondary optimization based on the temperature state data and the adjustment parameters to obtain the optimized temperature data, and input the optimized temperature data into the temperature controller for secondary regulation.

2. A multi-matrix device temperature cloud control method according to claim 1, characterized in that: The step of calibrating the real-time temperature to obtain the calibration temperature includes: Select a standard temperature position, and place a temperature sensor at the standard temperature position to measure the measured temperature of the standard temperature position, calculate the measured temperature and the standard temperature corresponding to the standard temperature position to obtain a calibration temperature difference; add the real-time temperature and the calibration temperature difference to obtain the calibration temperature.

3. A multi-matrix device temperature cloud control method according to claim 1, characterized in that: The device predicted temperature rise diagram is obtained based on the calibration temperature and the device operation data prediction, including: Extract calibration temperature and equipment operation data; wherein the equipment operation data includes equipment layout, load size, operation time and external environment temperature of multi-matrix equipment; Obtain the thermal impact data of multi-matrix equipment according to the equipment layout; define the unit time and the expected temperature that the equipment is expected to reach within the unit time; mark the temperature reached by the test equipment within the unit time as the test temperature, and mark the difference between the expected temperature and the test temperature as the error temperature; The unit impact rate WTS is calculated by the formula WTS=(CST-JZT)×WBT×α / (FZR×YXT); wherein CST represents the test temperature, JZT represents the calibration temperature, WBT represents the external ambient temperature, FZR represents the load size, YXT represents the operating time, and α represents the environmental impact coefficient, which is set based on the difference between the external ambient temperature and the test temperature; With time as the horizontal axis, the temperature of the equipment as the vertical axis, and the unit impact rate as the slope, a predicted temperature rise graph of the equipment is generated.

4. A multi-matrix device temperature cloud control method according to claim 3, characterized in that: The step of obtaining the thermal impact data of the multi-matrix equipment according to the equipment layout includes: Extracting device layout; wherein the device layout includes the length of cables connecting the multi-matrix devices and the distribution position of each multi-matrix device; Marking other multi-matrix devices connected to the multi-matrix device by data cables as connection devices, and marking the multi-matrix device to be subjected to temperature test as test device; measuring the connection device temperature of the connection device by a calibrated temperature sensor, and subtracting the connection device temperature from the calibration temperature to obtain a calibrated connection temperature; Determine whether the calibration temperature is greater than the calibration connection temperature; if yes, mark the thermal impact state of the connection device on the test device as a negative impact state; if no, mark the thermal impact state of the connection device on the test device as a positive impact state; Convert the distribution position of the multi-matrix device to obtain the distribution distance between the test device and the connected device; by the formula RYX = (JLT-JZT) × ln (FL) / ln (DL) 3 The layout temperature influence RYX is calculated; where JLT represents the calibration connection temperature, JZT represents the calibration temperature, FL represents the distribution distance, and DL represents the cable length; The layout temperature impact and thermal impact status are integrated to obtain the thermal impact data.

5. A multi-matrix device temperature cloud control method according to claim 3, characterized in that: The environmental impact factor is set based on the difference between the external ambient temperature and the test temperature, and includes: Extracting several groups of historical test data of the test equipment under known but different external ambient temperatures from the historical database; Randomly select a test running time, and mark the test temperature corresponding to the test running time in the historical test data as the historical test temperature; calculate the ratio of the difference between the historical test temperature and the calibration temperature and the difference between the external ambient temperature and the calibration temperature to obtain several groups of test influence coefficients; set the error term, and select the average of several groups of test influence coefficients as the verification coefficient; take the value obtained by subtracting the product of the difference between the external ambient temperature and the calibration temperature and the verification coefficient from the difference between the historical test temperature and the calibration temperature as the error value; Determine whether the error value corresponding to the historical test data meets the error term; if yes, retain the corresponding historical test data; if no, remove the corresponding historical test data; The mode of the test impact coefficients corresponding to the remaining historical test data is taken as the environmental impact coefficient.

6. A multi-matrix device temperature cloud control method according to claim 1, characterized in that: The adjustment parameters are obtained according to the comparison result between the equipment predicted temperature rise diagram and the standard temperature in the management database, including: Select the target time; obtain the predicted target temperature corresponding to the target time in the equipment predicted temperature rise diagram; Determine whether the predicted target temperature is greater than the standard temperature; if yes, mark the adjustment state as a cooling state; if no, mark the adjustment state as a warming state; The difference between the predicted target temperature and the standard temperature is marked as the adjustment temperature; the adjustment temperature and the adjustment state are integrated to obtain the adjustment parameter.

7. A multi-matrix device temperature cloud control method according to claim 1, characterized in that: The step of sending the adjustment parameters to the temperature controller for temperature control includes: Extract adjustment parameters; Determine whether the adjustment state is a cooling state; if yes, start the refrigeration element in the intelligent temperature control actuator; if not, start the heating element in the intelligent temperature control actuator; input the adjustment temperature into the temperature control model to obtain the temperature control action strategy, and input the temperature control action strategy into the intelligent temperature control actuator for temperature regulation; wherein, the temperature control model is constructed based on the neural network model.

8. A multi-matrix device temperature cloud control method according to claim 7, characterized in that: The temperature control model is constructed based on a neural network model, and includes: Several groups of historical adjustment parameters and corresponding historical temperature control action strategies are extracted from the historical database; the adjustment parameters and the corresponding historical temperature control action strategies are used as training data and test data, the neural network model is trained by the training data, the trained neural network model is tested by the test data, and the parameters of the neural network model are adjusted according to the test results to obtain a temperature control model with the input data as the adjustment parameters and the output data as the temperature control action strategy.

9. A multi-matrix device temperature cloud control method according to claim 1, characterized in that: The second optimization based on the temperature state data and the adjustment parameters to obtain the optimized temperature data includes: A1: Extract temperature status data and adjustment parameters; wherein the temperature status data includes the control temperature of the device after temperature control and the operation data of the control device at the corresponding time; A2: Mark the time corresponding to the controlled temperature as the controlled time, and match the predicted temperature corresponding to the controlled time in the device predicted temperature graph; A3: Determine whether the controlled temperature is the same as the predicted temperature; if yes, mark the corresponding temperature control state as normal; if no, mark the corresponding temperature control state as a state requiring optimization and jump to A4; A4: Determine whether the controlled temperature is lower than the predicted temperature; if yes, mark the optimized state as a state requiring heating; if no, mark the optimized state as a state requiring cooling; A5: Based on the temperature control state and the optimization state, the intelligent temperature control actuator is optimized twice to obtain the optimized temperature data.

10. A multi-matrix device temperature cloud control method according to claim 9, characterized in that: The second optimization of the intelligent temperature control actuator based on the temperature control state and the optimization state to obtain the optimized temperature data includes: For adjustment parameters whose temperature control state is a state that needs to be optimized, the environmental impact coefficient is updated according to the corresponding optimization state and temperature state parameters to obtain the updated impact coefficient; the updated temperature control action strategy and updated adjustment parameters are re-obtained through the updated influence coefficient, and the updated temperature control action strategy and updated adjustment parameters are integrated to obtain the optimized temperature data.

Images