A method for optimizing energy consumption of chiller units based on unsupervised learning algorithm

Abstract

The present invention discloses a method for optimizing energy consumption of a refrigeration unit based on an unsupervised learning algorithm, comprising: collecting operation and energy consumption parameters of equipment in the refrigeration unit; using the data in the refrigeration unit to perform real-time equipment energy consumption feature extraction using an unsupervised learning center, wherein the real-time equipment energy consumption feature extraction performed by the unsupervised learning center includes the operation status information of the refrigeration unit; using the unsupervised learning center to perform real-time equipment energy consumption optimization efficiency feature extraction; calculating whether the optimization efficiency of the real-time equipment energy consumption feature extraction performed by the unsupervised learning center is greater than a preset energy consumption threshold; when the optimization efficiency of the real-time equipment energy consumption feature extraction performed by the unsupervised learning center is calculated to be greater than the preset energy consumption threshold, the real-time equipment energy consumption feature extraction performed by the unsupervised learning center is calculated separately to the data control platform of each refrigeration unit module. The present invention can improve the accuracy of calculation and the optimization efficiency of refrigeration unit data.

Description

Technical Field

[0001] The present invention relates to the field of energy consumption control of refrigeration units, and in particular to a method for optimizing energy consumption of refrigeration units based on an unsupervised learning algorithm. Background Art

[0002] With the development of the times, industrial parks, large buildings, etc. now require refrigeration units to provide cold sources.

[0003] Chillers are used in the following applications: 1. Plastics Industry: Accurately controlling mold temperature during various plastic processing operations, shortening the molding cycle and ensuring consistent product quality. 2. Electronics Industry: Stabilizing the molecular structure of electronic components on the production line, improving the yield rate. They are also used in ultrasonic cleaning to effectively prevent the volatilization of expensive cleaning agents and the resulting damage. 3. Electroplating Industry: Controlling plating temperature increases the density and smoothness of plated parts, shortening plating cycles, improving production efficiency, and enhancing product quality. 4. Machinery Industry: Controlling oil pressure and temperature in hydraulic systems, stabilizing oil temperature and pressure, extending oil life, improving mechanical lubrication efficiency, and reducing wear. 5. Construction Industry: Supplying chilled water to concrete, ensuring the molecular structure of concrete is suitable for construction applications and effectively enhancing its hardness and toughness. 6. Vacuum Coating: Controlling the temperature of vacuum coating machines to ensure high-quality plated parts. 7. Food Industry: High-speed cooling of processed food to meet packaging requirements. Other applications include controlling the temperature of fermented foods. 8. Pharmaceutical Industry: Primarily used in the pharmaceutical industry to control the temperature of fermented pharmaceuticals. Pharmaceutical companies should make full use of chiller equipment.

[0004] Chillers consume a lot of power. Prolonged, heavy-load operation shortens their lifespan and increases costs for both the company and its users. Currently, load distribution for chillers is not intelligent enough, making it difficult to optimize energy consumption. Summary of the Invention

[0005] Based on this, it is necessary to address the problem of extracting real-time equipment energy consumption features from random unsupervised learning centers and provide a method for optimizing the energy consumption of refrigeration units based on unsupervised learning algorithms.

[0006] A method for optimizing energy consumption of a chiller unit based on an unsupervised learning algorithm, comprising:

[0007] Collect the operation and energy consumption parameters of the equipment in the refrigeration unit;

[0008] Using the operation and energy consumption parameters of the equipment in the refrigeration unit, an unsupervised learning center is used to extract real-time equipment energy consumption characteristics, wherein the real-time equipment energy consumption characteristics extracted by the unsupervised learning center include the operation status information of the refrigeration unit;

[0009] Utilizing the unsupervised learning center to extract efficiency features for real-time device energy consumption optimization;

[0010] Calculating whether the optimization efficiency of the unsupervised learning center for extracting real-time device energy consumption features is greater than a preset energy consumption threshold;

[0011] When the optimization efficiency of the unsupervised learning center for real-time equipment energy consumption feature extraction is greater than the preset energy consumption threshold, the unsupervised learning center used for real-time equipment energy consumption feature extraction is calculated separately to the data control platform of each refrigeration unit module; when the optimization efficiency of the unsupervised learning center for real-time equipment energy consumption feature extraction is not greater than the preset energy consumption threshold, the operation and energy consumption parameters of the equipment in the refrigeration unit are recalculated, the unsupervised learning center is used to perform real-time equipment energy consumption feature extraction and the optimization efficiency is calculated.

[0012] A method for optimizing energy consumption of a chiller unit based on an unsupervised learning algorithm, comprising:

[0013] The chiller unit operation status information construction module is used to collect the operation and energy consumption parameters of the equipment in the chiller unit;

[0014] A module for extracting real-time equipment energy consumption characteristics from an unsupervised learning center is connected to the chiller unit operation status information construction module and is used to extract real-time equipment energy consumption characteristics from the unsupervised learning center using the operation and energy consumption parameters of the equipment in the chiller unit. The real-time equipment energy consumption characteristics extracted by the unsupervised learning center include chiller unit operation status information;

[0015] an optimization efficiency module, connected to the module for extracting real-time device energy consumption features using the unsupervised learning center, and used to calculate the optimization efficiency of the unsupervised learning center for extracting real-time device energy consumption features used by the module for extracting real-time device energy consumption features using the unsupervised learning center;

[0016] an optimization efficiency management module, connected to the optimization efficiency module, for calculating whether the optimization efficiency of the unsupervised learning center for extracting real-time device energy consumption features is greater than a preset energy consumption threshold;

[0017] The data transmission module of the refrigeration unit module is connected to the module for extracting the real-time equipment energy consumption characteristics of the unsupervised learning center and the optimization efficiency management module. It is used to use the signal from the optimization efficiency management module indicating that the optimization efficiency of the real-time equipment energy consumption characteristics extraction of the unsupervised learning center is greater than the preset energy consumption threshold to calculate the real-time equipment energy consumption characteristics extraction of the unsupervised learning center used by the module for extracting the real-time equipment energy consumption characteristics of the unsupervised learning center to each refrigeration unit module data control platform.

[0018] The refrigeration unit module abnormal data optimization efficiency construction module is used to calculate the optimization efficiency of each refrigeration unit module abnormal data required for the optimization efficiency calculation of the unsupervised learning center for real-time equipment energy consumption feature extraction performed by the feature extraction optimization efficiency module, wherein the refrigeration unit module abnormal data optimization efficiency construction module is used to calculate the optimization efficiency and initial information of the refrigeration unit module abnormal data, and is used to calculate the secondary information of the refrigeration unit module abnormal data by using the probability of occurrence of the refrigeration unit module abnormal data in each unsupervised learning center used by the module for real-time equipment energy consumption feature extraction using the unsupervised learning center, and uses the optimization efficiency, initial information, and secondary information to calculate the new optimization efficiency of the refrigeration unit module abnormal data.

[0019] Beneficial effects:

[0020] The above-mentioned optimization method and optimization system for the unsupervised learning center to extract real-time equipment energy consumption characteristics calculate the optimization efficiency of the unsupervised learning center to extract real-time equipment energy consumption characteristics and compare it with the preset energy consumption threshold, thereby optimizing the real-time equipment energy consumption characteristics extraction of the unsupervised learning center by optimizing the optimization efficiency. Compared with the problems of prediction trouble and low accuracy brought about by existing random calculations, a refrigeration unit energy consumption optimization method based on an unsupervised learning algorithm in each embodiment of the present invention can optimize the calculation optimization difficulty, operability and horizontal differences of the real-time equipment energy consumption characteristics extraction of the unsupervised learning center by optimizing the optimization efficiency of the real-time equipment energy consumption characteristics extraction of the unsupervised learning center, thereby improving the usability of the refrigeration unit module data control platform, and improving the accuracy of calculation and the optimization efficiency of refrigeration unit data. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1 This is a first flow chart of a method for optimizing energy consumption of a chiller unit based on an unsupervised learning algorithm according to the present invention;

[0022] Figure 2 This is a second flow chart of a method for optimizing energy consumption of a chiller unit based on an unsupervised learning algorithm according to the present invention;

[0023] Figure 3 This is a third flow chart of a method for optimizing energy consumption of a chiller unit based on an unsupervised learning algorithm according to the present invention;

[0024] Figure 4 This is a module structure diagram of a refrigeration unit energy consumption optimization method based on an unsupervised learning algorithm of the present invention. DETAILED DESCRIPTION

[0025] It should be noted that, unless there is a conflict, the embodiments and features in the embodiments of this application can be combined with each other. The present application will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0026] like Figure 1 As shown, it is a process of a method for optimizing energy consumption of a refrigeration unit based on an unsupervised learning algorithm in one embodiment of the present invention.

[0027] Step 102, collect the operation and energy consumption parameters of the equipment in the refrigeration unit. For calculation optimization with specific content, the operation status information of the refrigeration unit is preferably the factor with the greatest influence on the calculation optimization calculated using the calculation optimization content. Generally, obtaining the operation status information of the refrigeration unit should be conducive to completing the abnormal data of the refrigeration unit module set according to the calculation optimization content. For example, for confrontational calculation optimization, the operation status information of the refrigeration unit can be set to the target feature or role with the greatest combat effectiveness. The operation and energy consumption parameters of the equipment in the refrigeration unit may include: the number, size, location, etc. of the operation status information of the refrigeration unit.

[0028] Step 104 : extracting real-time equipment energy consumption features using an unsupervised learning center using the operating and energy consumption parameters of the equipment in the refrigeration unit. The real-time equipment energy consumption features extracted by the unsupervised learning center include the operating status information of the refrigeration unit.

[0029] Specifically, the real-time equipment energy consumption feature extraction performed by the unsupervised learning center used should include the operating status information of the refrigeration unit. The real-time equipment energy consumption feature extraction performed by the unsupervised learning center should include the real-time equipment energy consumption feature extraction performed by the unsupervised learning center to be calculated to each refrigeration unit module data control platform using the calculation optimization content, wherein the refrigeration unit operating status information is included in the real-time equipment energy consumption feature extraction performed by the unsupervised learning center to be calculated to one or more refrigeration unit module data control platforms. For example, the refrigeration unit operating status information should be used when the real-time equipment energy consumption feature extraction is performed by the unsupervised learning center and included in the real-time equipment energy consumption feature extraction performed by the unsupervised learning center used.

[0030] Step 106 , calculating the optimization efficiency of the unsupervised learning center of the data control platform of each chiller module in performing real-time equipment energy consumption feature extraction.

[0031] Specifically, the unsupervised learning center to be calculated into the data control platform of each refrigeration unit module can be used to extract the real-time equipment energy consumption characteristics, and the optimization efficiency of the unsupervised learning center to be calculated into the data control platform of the refrigeration unit module for extracting the real-time equipment energy consumption characteristics can be calculated. It can be understood that the use of the unsupervised learning center generated by the random calculation method to extract the real-time equipment energy consumption characteristics will also have random differences in difficulty, operability, etc., so that unexpected unevenness may occur in the same calculation optimization and the same time period. This unexpected unevenness is also one of the important reasons why the existing random calculation principle affects the trouble and stickiness of user prediction. Using an embodiment of the present invention, calculating the optimization efficiency of the unsupervised learning center of the data control platform of each refrigeration unit module for extracting the real-time equipment energy consumption characteristics is an important means to optimize the existing random calculation principle.

[0032] Step 108 : Calculate the optimization efficiency of the unsupervised learning center for extracting real-time device energy consumption features.

[0033] Using one embodiment of the present invention, the optimization efficiency can be calculated based on the optimization efficiency of the unsupervised learning center of each chiller module data control platform performing real-time equipment energy consumption feature extraction calculated in step 106. Of course, using other feasible implementations, the optimization efficiency of the unsupervised learning center performing real-time equipment energy consumption feature extraction can also be calculated by performing preset operations on all used unsupervised learning centers. In this case, the calculation of the optimization efficiency of the unsupervised learning center performing real-time equipment energy consumption feature extraction on each chiller module data control platform in step 106 can also be omitted.

[0034] In this embodiment, the optimization efficiency reflects the difficulty and operability of the computational optimization content reflected in the unsupervised learning center's real-time equipment energy consumption feature extraction, as well as the differences between the unsupervised learning centers' real-time equipment energy consumption feature extraction for each chiller module data control platform. This optimization efficiency reveals the differences in the difficulty and operability of the unsupervised learning centers' real-time equipment energy consumption feature extraction for each chiller module data control platform, thereby distinguishing it from the unexpected difficulty, operability, and variability reflected in existing random calculation principles.

[0035] Step 110 , calculating whether the optimization efficiency of the unsupervised learning center for extracting real-time device energy consumption features is greater than a preset energy consumption threshold.

[0036] Specifically, the preset energy consumption threshold can be set based on the desired difference between the various computational optimization elements included in the real-time device energy consumption feature extraction using the unsupervised learning center and the real-time device energy consumption feature extraction using the unsupervised learning center of each chiller module data control platform. In this embodiment, when the optimization efficiency of the real-time device energy consumption feature extraction by the unsupervised learning center is not greater than the preset energy consumption threshold, it indicates that the difficulty and operability of the overall computational optimization and the difference between the real-time device energy consumption feature extraction by the unsupervised learning center of each chiller module data control platform are beyond expectations. Then, it is necessary to repeat the process of calculating the chiller unit operating status information, extracting the real-time device energy consumption feature using the unsupervised learning center, and calculating the optimization efficiency described in steps 102 to 108 until the optimization efficiency of the real-time device energy consumption feature extraction by the unsupervised learning center meets the expected difficulty, operability, and difference expectations, that is, the energy consumption threshold is greater than the preset energy consumption threshold.

[0037] Step 112: If the optimization efficiency of the unsupervised learning center for real-time equipment energy consumption feature extraction calculated in step 110 is greater than the preset energy consumption threshold, the unsupervised learning center used for real-time equipment energy consumption feature extraction is calculated separately to the data control platform of each refrigeration unit module.

[0038] In this embodiment, the optimization efficiency of the unsupervised learning center for real-time equipment energy consumption feature extraction reflects the overall difficulty, operability and individual differences of the unsupervised learning center for real-time equipment energy consumption feature extraction to be calculated for each refrigeration unit module data control platform. If the optimization efficiency is greater than the preset energy consumption threshold, it indicates that the overall difficulty, operability and individual differences of the unsupervised learning center for real-time equipment energy consumption feature extraction of each refrigeration unit module data control platform meet the required expectations, and each refrigeration unit module data control platform can use the unsupervised learning center for real-time equipment energy consumption feature extraction to start calculation optimization.

[0039] In an optional embodiment, before calculating the optimization efficiency of the unsupervised learning center for real-time equipment energy consumption feature extraction of each user in step 106, or before calculating the optimization efficiency of the unsupervised learning center for real-time equipment energy consumption feature extraction in step 108, the real-time equipment energy consumption feature extraction of the unsupervised learning center used can be calculated and sent to the data control platform of each chiller module. Therefore, if the optimization efficiency of the unsupervised learning center for real-time equipment energy consumption feature extraction is calculated to be no greater than a preset energy consumption threshold and it is necessary to recalculate the operation and energy consumption parameters of the equipment in the chiller unit and use the unsupervised learning center for real-time equipment energy consumption feature extraction, it is necessary to delete or withdraw the calculated unsupervised learning center for real-time equipment energy consumption feature extraction from the data control platform of each chiller module.

[0040] like Figure 2 As shown in FIG, it is a process of a method for optimizing energy consumption of a chiller unit based on an unsupervised learning algorithm according to another embodiment of the present invention. Figure 2 In, with Figure 1 Steps 202 to 212 and corresponding numbers in Figure 1 Steps 102 to 112 in the embodiment have similar contents and are not described in detail here.

[0041] Step 200 , regularly outputting a control signal, wherein the control signal indicates random optimization of the real-time device energy consumption feature extraction for the unsupervised learning center.

[0042] Specifically, the output of the control signal indicates that the present invention is different from the existing random computing unsupervised learning center in extracting real-time equipment energy consumption characteristics. That is, before the potential refrigeration unit module data control platform intends to enter the calculation optimization content, it will get a prompt about how the calculation optimization content is different from the existing random computing unsupervised learning center in extracting real-time equipment energy consumption characteristics.

[0043] Step 201: receiving control signal feedback and determining whether the signal feedback indicates acceptance of the random optimization of the real-time device energy consumption feature extraction of the unsupervised learning center.

[0044] Specifically, after the control signal of the aforementioned step 200 indicates that the computational optimization content to be started is different from the existing random computational optimization content, the potential refrigeration unit module data control platform can provide signal feedback based on the control signal, and the signal feedback indicates whether to accept the random optimization of the real-time equipment energy consumption feature extraction for the unsupervised learning center.

[0045] If the signal feedback indicates that the randomness optimization of the unsupervised learning center for extracting real-time device energy consumption features is accepted, then continue with step 202 and subsequent steps, use the unsupervised learning center to extract real-time device energy consumption features and perform calculations based on the optimized unsupervised learning center for extracting real-time device energy consumption features to start calculation optimization.

[0046] If the signal feedback indicates that the random optimization of the real-time equipment energy consumption feature extraction for the unsupervised learning center is not accepted, the calculation optimization content will not be opened to the potential calculation optimization user, and the process of regularly outputting the control signal and receiving and determining the control signal feedback in steps 200 and 201 will continue.

[0047] By utilizing the refrigeration unit energy consumption optimization method based on an unsupervised learning algorithm of this embodiment, it is possible to fully ensure that all parties using the calculation optimization understand the optimization of the random calculation method for extracting real-time equipment energy consumption characteristics of the existing unsupervised learning center by the refrigeration unit energy consumption optimization method of the present invention before the calculation optimization begins, and to ensure that the refrigeration unit module data control platform is fully informed before participating. In an optional embodiment, the notification information can be sent to the refrigeration unit module data control platform via a dialog box or prompt information, or can be reflected by setting a dedicated calculation optimization area, wherein in the dedicated calculated calculation optimization area, the information output of step 200 can be reflected in the form of an identification, prompt, etc. of the calculation optimization area; the received feedback information of step 201 can be reflected as whether the refrigeration unit module data control platform enters or does not enter the dedicated calculation optimization area.

[0048] like Figure 3 As shown, it is a process of a method for optimizing energy consumption of a chiller unit based on an unsupervised learning algorithm in another embodiment of the present invention. Figure 3 In, with Figure 1 Steps 302 to 312 and corresponding numbers in Figure 1 Steps 102 to 112 in the embodiment have similar contents and are not described in detail here.

[0049] In this embodiment, after calculating in step 310 whether the optimization efficiency of the unsupervised learning center for extracting real-time equipment energy consumption characteristics is greater than the preset energy consumption threshold, if the optimization efficiency of the unsupervised learning center for extracting real-time equipment energy consumption characteristics is greater than the preset energy consumption threshold, step 311 is performed to determine whether the number of times the unsupervised learning center used to extract real-time equipment energy consumption characteristics is calculated to each refrigeration unit module data control platform under the energy consumption threshold has reached a preset number parameter.

[0050] If the calculation in step 311 shows that the number of times the unsupervised learning center performs real-time equipment energy consumption feature extraction and calculation on each refrigeration unit module data control platform under the energy consumption threshold has not reached the preset number parameter, then continue to step 312 and perform real-time equipment energy consumption feature extraction and calculation on each refrigeration unit module data control platform using the unsupervised learning center.

[0051] In step 314, if the number of times calculated in step 311 has reached the preset number parameter, a new energy consumption threshold is calculated, and then step 310 is performed to compare the optimization efficiency of the unsupervised learning center for extracting real-time device energy consumption features with the new energy consumption threshold.

[0052] Through the refrigeration unit energy consumption optimization method based on an unsupervised learning algorithm of this embodiment, dynamic adjustment of the energy consumption threshold for the unsupervised learning center for real-time equipment energy consumption feature extraction can be achieved. Optionally, since the refrigeration unit energy consumption optimization method based on an unsupervised learning algorithm of the embodiment of the present invention only optimizes the unsupervised learning center for real-time equipment energy consumption feature extraction in terms of refrigeration unit operating status information and optimization efficiency, the dynamic optimization of the energy consumption threshold can utilize the randomness of the unsupervised learning center for real-time equipment energy consumption feature extraction in addition to the refrigeration unit operating status information to optimize the real-time equipment energy consumption feature extraction of the unsupervised learning center, thereby further optimizing the randomness.

[0053] Another embodiment of the present invention is a process of a method for optimizing energy consumption of a chiller unit based on an unsupervised learning algorithm. Figure 1 Steps 402 to 412 and corresponding numbers in Figure 1 Steps 102 to 112 in the embodiment have similar contents and are not described in detail here.

[0054] In this embodiment, if the optimization efficiency of the unsupervised learning center to be calculated for real-time equipment energy consumption feature extraction calculated in step 410 is not greater than the energy consumption threshold and it is necessary to recalculate the refrigeration unit operating status information, use the unsupervised learning center to perform real-time equipment energy consumption feature extraction, and calculate the optimization efficiency, step 414 archives the current unsupervised learning center for real-time equipment energy consumption feature extraction and its corresponding optimization efficiencies of all parties.

[0055] Step 416 , calculate whether the number of repetitions of repeatedly performing the cooling unit operating status information, extracting the real-time equipment energy consumption characteristics using the unsupervised learning center, and calculating the optimization efficiency under the current optimization efficiency and energy consumption threshold comparison round reaches a preset repetition parameter.

[0056] Here, the current round of comparison of optimization efficiency and energy consumption threshold refers to the comparison of optimization efficiency and energy consumption threshold of real-time device energy consumption feature extraction performed by the unsupervised learning center in step 410 since the last calculation of real-time device energy consumption feature extraction by the unsupervised learning center.

[0057] Specifically, after using the operating and energy consumption parameters of the equipment in the refrigeration unit and using an unsupervised learning center that includes the operating status information of the refrigeration unit to extract real-time equipment energy consumption characteristics, if the optimization efficiency is not greater than a preset energy consumption threshold, it is necessary to re-collect the operating and energy consumption parameters of the equipment in the refrigeration unit, and then recalculate the unsupervised learning center to extract real-time equipment energy consumption characteristics to calculate the corresponding optimization efficiency. However, it is possible that after recalculating the operating status information of the refrigeration unit and extracting the real-time equipment energy consumption characteristics of the unsupervised learning center several times in a row, the obtained optimization efficiency is still not greater than the preset energy consumption threshold. In order to save the time required to recalculate the operating status information of the refrigeration unit and extract the real-time equipment energy consumption characteristics of the unsupervised learning center, it is necessary to limit the number of such repetitions.

[0058] If the number of repetitions does not reach the preset repetition parameter, the process of calculating the operating status information of the refrigeration unit, extracting the real-time equipment energy consumption characteristics and optimizing the efficiency thereof by the unsupervised learning center, and comparing with the preset energy consumption threshold value as described in steps 402 to 410 is continued.

[0059] In step 418, if the number of repetitions calculated in step 416 has reached the preset repetition parameter under the comparison round of the current optimization efficiency and the energy consumption threshold, the unsupervised learning center with the largest optimization efficiency archived is calculated to extract the real-time equipment energy consumption characteristics as the unsupervised learning center to be calculated, and step 412 is continued to calculate the real-time equipment energy consumption characteristics of the unsupervised learning center to be calculated to the data control platform of each refrigeration unit module to start calculation optimization.

[0060] Using this embodiment of the present invention, in order to avoid the optimization of the unsupervised learning center for real-time equipment energy consumption feature extraction being too lengthy and affecting the user experience, it is necessary to limit the number of times the unsupervised learning center is repeatedly used for real-time equipment energy consumption feature extraction. If the number of comparisons reaching the repetition parameter has been performed using step 410 since the last calculation of the unsupervised learning center for real-time equipment energy consumption feature extraction to each refrigeration unit module data control platform, then in order to avoid further calculating the refrigeration unit operating status information and the time spent on the process of using the unsupervised learning center for real-time equipment energy consumption feature extraction, it is necessary to use the unsupervised learning center for real-time equipment energy consumption feature extraction with the highest optimization efficiency among the unsupervised learning center for real-time equipment energy consumption feature extraction used after the last calculation of the unsupervised learning center for real-time equipment energy consumption feature extraction as the unsupervised learning center for real-time equipment energy consumption feature extraction to be calculated and calculated to each refrigeration unit module data control platform using the process of step 412.

[0061] By utilizing the above embodiments of the present invention, the optimization efficiency of the unsupervised learning center for real-time equipment energy consumption feature extraction is calculated and compared with the preset energy consumption threshold, thereby optimizing the optimization efficiency of the unsupervised learning center for real-time equipment energy consumption feature extraction. Compared with the existing random calculation of the unsupervised learning center for real-time equipment energy consumption feature extraction, which brings about the prediction troubles and low accuracy problems, a refrigeration unit energy consumption optimization method based on an unsupervised learning algorithm in each embodiment of the present invention can optimize the calculation optimization difficulty, operability and horizontal differences of the unsupervised learning center for real-time equipment energy consumption feature extraction by optimizing the optimization efficiency of the unsupervised learning center, thereby improving the usability of the refrigeration unit module data control platform and improving the accuracy of the calculation.

[0062] In another embodiment of the present invention, in an optimization method for extracting real-time equipment energy consumption characteristics by an unsupervised learning center, a process for calculating the optimization efficiency of the unsupervised learning center of each refrigeration unit module data control platform for extracting real-time equipment energy consumption characteristics is provided.

[0063] Step 502: Calculate the optimized efficiency of the chiller unit operating status information.

[0064] Specifically, in Figure 1 After steps 102, 202, 302, and 402 collect the operating and energy consumption parameters of the equipment in the chiller unit, and steps 104, 204, 304, and 404 use the operating and energy consumption parameters of the equipment in the chiller unit to calculate the unsupervised learning center that includes the chiller unit operating status information to perform real-time equipment energy consumption feature extraction, the real-time equipment energy consumption feature extraction to be calculated by the unsupervised learning center of the data control platform of each chiller unit module may or may not include the chiller unit operating status information. Therefore, in step 502, if the real-time equipment energy consumption feature extraction performed by the unsupervised learning center of the chiller unit module data control platform includes the chiller unit operating status information, the optimization efficiency of the chiller unit operating status information is calculated. If the chiller unit operating status information is not included, the process proceeds to the next step to calculate the optimization efficiency of the next chiller unit module abnormal data. Similarly, it can be understood that in the process of calculating the optimization efficiency of abnormal data of other refrigeration unit modules, it is only calculated when the abnormal data of the refrigeration unit module is present in the real-time equipment energy consumption feature extraction of the unsupervised learning center of the refrigeration unit module data control platform.

[0065] Step 504 : Calculate the optimization efficiency of the abnormal data of the re-cooling unit module.

[0066] As previously mentioned, the chiller unit operating status information is typically the chiller unit module abnormality data that has a significant impact on the overall optimization progress and operation. Furthermore, after the chiller unit operating status information, the optimization efficiency of the chiller unit module abnormality data, which has a slightly less significant impact, can also be sequentially calculated.

[0067] Step 506: Calculate the final optimization efficiency of the abnormal data of the chiller module.

[0068] Generally speaking, in the unsupervised learning center of the refrigeration unit module data control platform to be calculated for real-time equipment energy consumption feature extraction, a variety of refrigeration unit module abnormal data can be included, and the optimization efficiency of various refrigeration unit module abnormal data can be calculated in turn according to the influence of various types of refrigeration unit module abnormal data.

[0069] Step 508 : Utilizing the calculated optimization efficiency of the abnormal data of each chiller module, the optimization efficiency of the unsupervised learning center of the chiller module data control platform for extracting real-time equipment energy consumption features is calculated.

[0070] Specifically, after calculating the optimization efficiency of various types of abnormal data for chiller modules, the optimization efficiency of the unsupervised learning center of the chiller module data control platform for real-time equipment energy consumption feature extraction can be calculated according to a preset method. In an optional embodiment, the optimization efficiency of the unsupervised learning center of a chiller module data control platform for real-time equipment energy consumption feature extraction can be the sum or weighted sum of the optimization efficiencies of all abnormal data for chiller modules included in the real-time equipment energy consumption feature extraction performed by the unsupervised learning center.

[0071] The above steps 502 to 506 only illustrate the process of calculating the optimization efficiency of three types of refrigeration unit module abnormality data of the refrigeration unit module data control platform. It should be understood that for a calculation optimization content, more or fewer refrigeration unit module abnormality data may be included. In this case, the optimization efficiency of each type of refrigeration unit module abnormality data can be similarly calculated, and the optimization efficiency of the unsupervised learning center of the refrigeration unit module data control platform for real-time equipment energy consumption feature extraction can be calculated according to the process shown in step 508. In addition, the process of calculating the optimization efficiency of various refrigeration unit module abnormality data can also be calculated in a different order than the order of the influence of the refrigeration unit module abnormality data as described above. Any other order is also feasible.

[0072] Furthermore, as mentioned above, Figure 1The process of calculating the optimization efficiency of the unsupervised learning center for real-time equipment energy consumption feature extraction described in steps 108, 208, 308, and 408 can be calculated using the optimization efficiency of the unsupervised learning center for real-time equipment energy consumption feature extraction of each refrigeration unit module data control platform. It can also be calculated by performing preset operations on all used unsupervised learning centers for real-time equipment energy consumption feature extraction. As an optional embodiment, the optimization efficiency of the unsupervised learning center for real-time equipment energy consumption feature extraction can be an algorithm for the optimization efficiency of the unsupervised learning center for real-time equipment energy consumption feature extraction of each refrigeration unit module data control platform and the optimization efficiency difference.

[0073] It reflects the differences between the real-time equipment energy consumption feature extraction performed by the unsupervised learning center calculated by the data control platform of each refrigeration unit module.

[0074] Utilizing a further optional embodiment of the present invention, the optimization efficiency of each type of refrigeration unit module abnormality data is not fixed, but can be dynamically adjusted. A process for dynamically adjusting the optimization efficiency of refrigeration unit module abnormality data utilizing one embodiment of the present invention. It should be understood that the dynamic adjustment of the optimization efficiency of refrigeration unit module abnormality data in this embodiment can be applied to each of multiple types of refrigeration unit module abnormality data, and other possible dynamic adjustment methods are also applicable.

[0075] Step 602: Calculate the optimized efficiency and initial information of abnormal data of the chiller module.

[0076] Specifically, for a calculation optimization content of a calculation, a predetermined frequency of occurrence of abnormal data of the refrigeration unit module is calculated, and the expected occurrence ratio is used as initial information of the abnormal data of the refrigeration unit module.

[0077] Step 604 : After the unsupervised learning center performs calculations for extracting the real-time equipment energy consumption characteristics for a preset number of times, the abnormal data of the chiller module is calculated again.

[0078] It is understandable that although, as mentioned above, the predetermined frequency of occurrence of abnormal data of the refrigeration unit module is calculated, in the limited number of unsupervised learning centers for real-time equipment energy consumption feature extraction and calculation, the occurrence of abnormal data of the refrigeration unit module may not completely comply with the predetermined frequency of occurrence. In the calculation of real-time equipment energy consumption feature extraction by the unsupervised learning center for the preset number of times, the probability of occurrence of the abnormal data of the refrigeration unit module is calculated as the repeated information of the abnormal data of the refrigeration unit module.

[0079] Step 606 : Calculate a new optimized efficiency of the abnormal data of the chiller module by using the optimized efficiency, the initial information, and the secondary information.

[0080] Therefore, it can be understood that when performing real-time equipment energy consumption feature extraction calculations in the unsupervised learning center for the preset number of times, if the probability of occurrence of abnormal data of the refrigeration unit module is greater than expected, the optimization efficiency of the abnormal data of the refrigeration unit module is lowered.

[0081] Furthermore, if dynamic adjustment of the optimization efficiency of chiller module abnormality data has already been performed during the calculation optimization process, the optimization efficiency and occurrence probability of chiller module abnormality data from the previous preset number of unsupervised learning center real-time equipment energy consumption feature extraction calculations will be used as the benchmark ratio for the optimization efficiency and occurrence of chiller module abnormality data in the new dynamic adjustment. In other words, the optimization efficiency and secondary information from the previous dynamic adjustment will be used as the optimization efficiency and initial information for the next dynamic adjustment.

[0082] By dynamically adjusting the optimization efficiency of the abnormal data of the refrigeration unit module in this embodiment, it is possible to avoid affecting the operability due to the randomness of the abnormal data of the refrigeration unit module during the real-time equipment energy consumption feature extraction and calculation process in the actual unsupervised learning center, thereby further realizing the optimization and optimization of randomness.

[0083] Utilize the further optional embodiment of the present invention, in the aforementioned Figure 1 In the process of dynamically adjusting the energy consumption thresholds in steps 110, 210, 310, and 410, the energy consumption thresholds compared are not fixed but can be adjusted dynamically. It should be understood that other possible dynamic adjustment methods are also applicable.

[0084] Step 702: Calculate the initial energy consumption threshold.

[0085] Specifically, before the calculation optimization starts, the initial energy consumption threshold is a default initial value set.

[0086] Step 704 : After performing the calculation of the real-time device energy consumption feature extraction by the unsupervised learning center for a preset number of times, the average value of the optimization efficiency of the real-time device energy consumption feature extraction by the unsupervised learning center in each calculation is calculated.

[0087] Specifically, in the actual calculation process of the unsupervised learning center for real-time device energy consumption feature extraction, in fact, the distribution of the optimization efficiency of the unsupervised learning center for real-time device energy consumption feature extraction that meets the condition of being greater than the energy consumption threshold may be uneven. Taking the average value of the optimization efficiency of the unsupervised learning center for real-time device energy consumption feature extraction in the preset number of unsupervised learning centers for real-time device energy consumption feature extraction calculations can reflect the calculation status of the optimization efficiency in the unsupervised learning center for real-time device energy consumption feature extraction calculations for the preset number of times.

[0088] In step 706 , a new energy consumption threshold is calculated based on the average of the initial energy consumption threshold and the optimized efficiency.

[0089] Furthermore, if the energy consumption threshold has already been dynamically adjusted during the optimization process, the energy consumption threshold used in the unsupervised learning center's previous preset number of real-time device energy consumption feature extraction calculations will be used as the initial energy consumption threshold for the new dynamic adjustment. By dynamically adjusting the parameters, it is possible to prevent the unsupervised learning center's real-time device energy consumption feature extraction calculations from being affected by setting the parameters too high or too low.

[0090] like Figure 4 As shown, it is a structural and functional diagram of an unsupervised learning center for real-time equipment energy consumption feature extraction and optimization system according to an embodiment of the present invention.

[0091] The unsupervised learning center performs real-time equipment energy consumption feature extraction and optimization system, which includes:

[0092] The chiller unit operation status information construction module is used to collect the operation and energy consumption parameters of the equipment in the chiller unit.

[0093] For computational optimization with specific content, the chiller unit operating status information is preferably the factor with the greatest influence on the computational optimization, calculated using the computational optimization content. Generally, obtaining the chiller unit operating status information should facilitate completing the chiller unit module abnormality data set based on the computational optimization content. For example, for adversarial computational optimization, the chiller unit operating status information can be set to the target feature or role with the greatest combat effectiveness. The operating and energy consumption parameters of the equipment in the chiller unit may include: the number, size, and location of the chiller unit operating status information.

[0094] The unsupervised learning center performs real-time equipment energy consumption feature extraction using a module, which is used to use the operation and energy consumption parameters of the equipment in the refrigeration unit to perform real-time equipment energy consumption feature extraction using the unsupervised learning center. The real-time equipment energy consumption feature extraction performed by the unsupervised learning center includes the operation status information of the refrigeration unit.

[0095] Specifically, the unsupervised learning center used by the module for extracting real-time equipment energy consumption features should include the operating status information of the refrigeration unit. The unsupervised learning center used by the module for extracting real-time equipment energy consumption features should include the unsupervised learning center used by the module for extracting real-time equipment energy consumption features to be calculated to each refrigeration unit module data control platform using the calculation optimization content, wherein the refrigeration unit operating status information is included in the unsupervised learning center used by the module for extracting real-time equipment energy consumption features to be calculated to one or more refrigeration unit module data control platforms. For example, the refrigeration unit operating status information should be used and included in the unsupervised learning center used by the module for extracting real-time equipment energy consumption features to be used by the module for extracting real-time equipment energy consumption features to be used by the module for extracting real-time equipment energy consumption features to be used by the module.

[0096] The optimization efficiency module is used to calculate the optimization efficiency of the unsupervised learning center of the data control platform of each refrigeration unit module for real-time equipment energy consumption feature extraction, and to calculate the optimization efficiency of the unsupervised learning center for real-time equipment energy consumption feature extraction.

[0097] Specifically, the optimization efficiency module can utilize an unsupervised learning center to extract real-time equipment energy consumption characteristics. The module uses the unsupervised learning center to be calculated into the data control platform of each chiller module to extract real-time equipment energy consumption characteristics, and calculates the optimization efficiency of the unsupervised learning center to be calculated into the data control platform of the chiller module for real-time equipment energy consumption characteristics extraction. It is understandable that the unsupervised learning center generated by a random calculation method for real-time equipment energy consumption characteristics extraction also has random differences in difficulty, operability, and other aspects, resulting in unexpected unevenness within the same calculation optimization period. This unexpected unevenness is also one of the important reasons why existing random calculation principles affect user prediction troubles and stickiness. Using one embodiment of the present invention, using the optimization efficiency module to calculate the optimization efficiency of the unsupervised learning center of each chiller module data control platform for real-time equipment energy consumption characteristics extraction is an important means of optimizing existing random calculation principles.

[0098] By using one embodiment of the present invention, the optimization efficiency module can calculate the optimization efficiency of the unsupervised learning center for real-time equipment energy consumption feature extraction based on the calculated optimization efficiency of the unsupervised learning center for real-time equipment energy consumption feature extraction of each refrigeration unit module data control platform. Of course, by using other feasible implementation methods, it is also possible to calculate the optimization efficiency of the unsupervised learning center for real-time equipment energy consumption feature extraction by performing preset operations on all unsupervised learning centers used by the module for real-time equipment energy consumption feature extraction of the unsupervised learning center. In this case, the optimization efficiency module may not calculate the optimization efficiency of the unsupervised learning center for real-time equipment energy consumption feature extraction of each refrigeration unit module data control platform.

[0099] In this embodiment, the optimization efficiency reflects the difficulty and operability of the computational optimization content reflected in the unsupervised learning center's real-time equipment energy consumption feature extraction, as well as the differences between the unsupervised learning centers' real-time equipment energy consumption feature extraction for each chiller module data control platform. This optimization efficiency reveals the differences in the difficulty and operability of the unsupervised learning centers' real-time equipment energy consumption feature extraction for each chiller module data control platform, thereby distinguishing it from the unexpected difficulty, operability, and variability reflected in existing random calculation principles.

[0100] The optimization efficiency management module is used to calculate whether the optimization efficiency of the unsupervised learning center for real-time equipment energy consumption feature extraction is greater than a preset energy consumption threshold.

[0101] Specifically, the preset energy consumption threshold can be set based on the desired difference between the various computational optimization elements included in the real-time device energy consumption feature extraction using the unsupervised learning center and the real-time device energy consumption feature extraction using the unsupervised learning center of each chiller module data control platform. In this embodiment, when the optimization efficiency management module calculates the optimization efficiency of the unsupervised learning center for real-time device energy consumption feature extraction is not greater than the preset energy consumption threshold, indicating that the overall computational optimization difficulty and operability, as well as the difference between the real-time device energy consumption feature extraction of the unsupervised learning center of each chiller module data control platform, are beyond expectations, it is necessary to notify the chiller unit operation status information construction module, the unsupervised learning center for real-time device energy consumption feature extraction module, and the optimization efficiency module to respectively repeat the aforementioned processes of calculating chiller unit operation status information, extracting real-time device energy consumption features using the unsupervised learning center, and calculating the optimization efficiency, until the optimization efficiency calculated by the optimization efficiency management module for real-time device energy consumption feature extraction using the unsupervised learning center meets the expected difficulty, operability, and difference expectations, that is, the energy consumption threshold is greater than the preset energy consumption threshold.

[0102] The data transmission module of the refrigeration unit module is connected to the module for extracting the real-time equipment energy consumption characteristics of the unsupervised learning center and the optimization efficiency management module. It is used to use the signal from the optimization efficiency management module indicating that the optimization efficiency of the real-time equipment energy consumption characteristics extraction of the unsupervised learning center is greater than the preset energy consumption threshold to calculate the real-time equipment energy consumption characteristics extraction of the unsupervised learning center used by the module for extracting the real-time equipment energy consumption characteristics of the unsupervised learning center to each refrigeration unit module data control platform.

[0103] The refrigeration unit module abnormal data optimization efficiency construction module is used to calculate the optimization efficiency of each refrigeration unit module abnormal data required for the optimization efficiency calculation of the unsupervised learning center for real-time equipment energy consumption feature extraction performed by the feature extraction optimization efficiency module, wherein the refrigeration unit module abnormal data optimization efficiency construction module is used to calculate the optimization efficiency and initial information of the refrigeration unit module abnormal data, and is used to calculate the secondary information of the refrigeration unit module abnormal data by using the probability of occurrence of the refrigeration unit module abnormal data in each unsupervised learning center used by the module for real-time equipment energy consumption feature extraction using the unsupervised learning center, and uses the optimization efficiency, initial information, and secondary information to calculate the new optimization efficiency of the refrigeration unit module abnormal data.

[0104] In this embodiment, the optimization efficiency of the unsupervised learning center for real-time equipment energy consumption feature extraction reflects the overall difficulty, operability and individual differences of the unsupervised learning center for real-time equipment energy consumption feature extraction to be calculated for each refrigeration unit module data control platform. If the optimization efficiency is greater than the preset energy consumption threshold, it indicates that the overall difficulty, operability and individual differences of the unsupervised learning center for real-time equipment energy consumption feature extraction of each refrigeration unit module data control platform meet the required expectations, and each refrigeration unit module data control platform can use the unsupervised learning center for real-time equipment energy consumption feature extraction to start calculation optimization.

[0105] In an optional embodiment, the system may further include a register connected to the module for extracting real-time device energy consumption characteristics from the unsupervised learning center, and used to store the unsupervised learning center's real-time device energy consumption characteristics used by the module. Thus, when the optimization efficiency management module calculates that the optimization efficiency of the unsupervised learning center's real-time device energy consumption characteristics is greater than a preset energy consumption threshold, the chiller unit module data transmission module may retrieve the stored unsupervised learning center's real-time device energy consumption characteristics from the register and send the unsupervised learning center's real-time device energy consumption characteristics to the data control platform of each chiller unit module.

[0106] In an optional embodiment, the refrigeration unit module data transmission module can directly calculate the unsupervised learning center used by the module for real-time equipment energy consumption feature extraction or the unsupervised learning center stored in the register for real-time equipment energy consumption feature extraction to each refrigeration unit module data control platform without relying on the instruction signal of the optimization efficiency management module. Thus, when the optimization efficiency management module calculates that the optimization efficiency of the unsupervised learning center for real-time equipment energy consumption feature extraction is not greater than the preset energy consumption threshold and it is necessary to recalculate the operation and energy consumption parameters of the equipment in the refrigeration unit and use the unsupervised learning center for real-time equipment energy consumption feature extraction, the refrigeration unit module data transmission module needs to send a command to each refrigeration unit module data control platform to delete or withdraw the calculated unsupervised learning center for real-time equipment energy consumption feature extraction from each refrigeration unit module data control platform.

[0107] By utilizing the above embodiments of the present invention, the optimization efficiency of the unsupervised learning center for real-time equipment energy consumption feature extraction is calculated and compared with the preset energy consumption threshold, thereby optimizing the optimization efficiency of the unsupervised learning center for real-time equipment energy consumption feature extraction. Compared with the existing random calculation of the unsupervised learning center for real-time equipment energy consumption feature extraction, the unsupervised learning center for real-time equipment energy consumption feature extraction optimization system of each embodiment of the present invention can optimize the calculation optimization difficulty, operability and horizontal differences of the unsupervised learning center for real-time equipment energy consumption feature extraction by optimizing the optimization efficiency of the unsupervised learning center, thereby improving the usability of the refrigeration unit module data control platform and improving the accuracy of the calculation.

[0108] In the description of the present invention, it should be noted that, unless otherwise specified or limited, the terms "disposed," "installed," "connected," "connected," and "fixed" should be understood in a broad sense. For example, they can refer to fixed connections, detachable connections, or integral connections; they can refer to mechanical connections or electrical connections; they can refer to direct connections or indirect connections through an intermediate medium; and they can refer to internal communication between two components. Those skilled in the art will understand the specific meanings of the above terms in the present invention based on specific circumstances.

[0109] Although embodiments of the present invention have been shown and described, it will be understood by those skilled in the art that various equivalent changes, modifications, substitutions and variations may be made to these embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for optimizing energy consumption of a chiller unit based on an unsupervised learning algorithm, characterized in that: include: Collect the operation and energy consumption parameters of the equipment in the refrigeration unit; Using the operation and energy consumption parameters of the equipment in the refrigeration unit, an unsupervised learning center is used to extract real-time equipment energy consumption characteristics, wherein the real-time equipment energy consumption characteristics extracted by the unsupervised learning center include the operation status information of the refrigeration unit; Utilizing the unsupervised learning center to extract efficiency features for real-time device energy consumption optimization; Calculating whether the optimization efficiency of the unsupervised learning center for extracting real-time device energy consumption features is greater than a preset energy consumption threshold; When the optimization efficiency of the unsupervised learning center for extracting the real-time equipment energy consumption characteristics is greater than the preset energy consumption threshold, the unsupervised learning center used for extracting the real-time equipment energy consumption characteristics is calculated and transmitted to the data control platform of each chiller module; When the optimization efficiency of the unsupervised learning center for extracting the real-time equipment energy consumption characteristics is calculated to be no greater than a preset energy consumption threshold, recalculating the operation and energy consumption parameters of the equipment in the refrigeration unit, reusing the unsupervised learning center to extract the real-time equipment energy consumption characteristics, and recalculating the optimization efficiency; The steps of extracting real-time device energy consumption optimization efficiency features using the unsupervised learning center include: Calculate the optimization efficiency of the unsupervised learning center of the data control platform of each chiller module to extract the real-time equipment energy consumption characteristics; The optimization efficiency of the unsupervised learning center of the data control platform of each chiller module in extracting the real-time equipment energy consumption characteristics and the algorithm of the optimization efficiency difference are used as the optimization efficiency of the unsupervised learning center in extracting the real-time equipment energy consumption characteristics; Before the step of extracting the real-time equipment energy consumption characteristics of the unsupervised learning center and calculating them separately into the data control platform of each chiller module, the corresponding optimized efficiency and initial information of the abnormal data of each chiller module included in the real-time equipment energy consumption characteristics extraction by the unsupervised learning center are calculated, and the initial information is the predetermined frequency of occurrence of the abnormal data of the corresponding chiller module; After performing a preset number of calculations for extracting the real-time equipment energy consumption characteristics by the unsupervised learning center, calculating corresponding secondary information of abnormal data of each chiller module, the secondary information being the probability of occurrence of the abnormal data of the corresponding chiller module in the calculations for extracting the real-time equipment energy consumption characteristics by the unsupervised learning center for the preset number of times; Calculating a new optimized efficiency of the abnormal data of the chiller module by using the optimized efficiency, the initial information and the secondary information; The unsupervised learning center extracts real-time equipment energy consumption features including abnormal data of each refrigeration unit module, and the optimization efficiency of the unsupervised learning center extracts real-time equipment energy consumption features is calculated using the optimization efficiency of the abnormal data of each refrigeration unit module.

2. The method for optimizing energy consumption of a chiller unit based on an unsupervised learning algorithm according to claim 1, characterized in that: When the optimization efficiency of the unsupervised learning center for extracting real-time device energy consumption features is calculated to be no greater than a preset energy consumption threshold, the method further includes: Archiving the current unsupervised learning center to extract the real-time equipment energy consumption characteristics and the corresponding unsupervised learning center of each chiller module data control platform to extract the optimization efficiency of the real-time equipment energy consumption characteristics; calculating whether the number of repetitions of calculating the operation and energy consumption parameters of the equipment in the chiller unit, extracting the real-time equipment energy consumption characteristics using the unsupervised learning center, and calculating the optimization efficiency reaches a preset repetition parameter since the last time the unsupervised learning center performed the calculation for extracting the real-time equipment energy consumption characteristics; When the number of repetitions reaches a preset repetition parameter, the unsupervised learning center with the largest optimization efficiency is calculated and used as the unsupervised learning center to be calculated for real-time device energy consumption feature extraction; When the calculated number of repetitions does not reach the preset repetition parameter, the operation and energy consumption parameter calculation of the equipment in the refrigeration unit is performed again, and the real-time equipment energy consumption feature extraction and calculation optimization efficiency are performed using the unsupervised learning center.

3. The method for optimizing energy consumption of a chiller unit based on an unsupervised learning algorithm according to claim 1, characterized in that: After the optimization efficiency of the real-time device energy consumption feature extraction in the calculation unsupervised learning center is greater than a preset energy consumption threshold, the method further includes: Determine whether the number of times the unsupervised learning center used performs real-time equipment energy consumption feature extraction and calculation on each chiller module data control platform has reached a preset number parameter under the energy consumption threshold; When the number of times calculated has reached a preset number parameter, a new energy consumption threshold is calculated.

4. The method for optimizing energy consumption of a chiller unit based on an unsupervised learning algorithm according to claim 3 is characterized in that: Calculating a new energy consumption threshold includes: Calculate the initial energy consumption threshold; Calculate the average value of the optimization efficiency of the unsupervised learning center for real-time equipment energy consumption feature extraction in each calculation; A new energy consumption threshold is calculated according to the average value of the initial energy consumption threshold and the optimized efficiency, wherein the new energy consumption threshold is the average value of the initial energy consumption threshold and the optimized efficiency.

5. A system utilizing the method for optimizing energy consumption of a chiller unit based on an unsupervised learning algorithm according to claim 1, characterized in that: include: The chiller unit operation status information construction module is used to collect the operation and energy consumption parameters of the equipment in the chiller unit; An unsupervised learning center performs real-time equipment energy consumption feature extraction using a module, configured to utilize the operation and energy consumption parameters of the equipment in the refrigeration unit to perform real-time equipment energy consumption feature extraction using the unsupervised learning center, wherein the real-time equipment energy consumption feature extraction performed by the unsupervised learning center includes refrigeration unit operation status information; An optimization efficiency module, used to calculate the optimization efficiency of the unsupervised learning center used by the module for extracting the real-time device energy consumption features; An optimization efficiency management module is used to calculate whether the optimization efficiency of the unsupervised learning center for extracting real-time device energy consumption features is greater than a preset energy consumption threshold; The chiller module data transmission module is used to utilize the signal from the optimization efficiency management module indicating that the optimization efficiency of the unsupervised learning center for real-time equipment energy consumption feature extraction is greater than a preset energy consumption threshold, and calculate the real-time equipment energy consumption feature extraction of the unsupervised learning center used by the module for real-time equipment energy consumption feature extraction to the data control platform of each chiller module; A module for optimizing the efficiency of abnormal data of a refrigeration unit module is used to calculate the optimization efficiency of each abnormal data of the refrigeration unit module required for optimizing the efficiency of the unsupervised learning center for real-time equipment energy consumption feature extraction performed by the feature extraction optimization efficiency module, wherein the module for optimizing the efficiency of abnormal data of the refrigeration unit module is used to calculate the optimization efficiency and initial information of the abnormal data of the refrigeration unit module, and is used to calculate the secondary information of the abnormal data of the refrigeration unit module based on the probability of occurrence of the abnormal data of the refrigeration unit module in each unsupervised learning center used by the module for real-time equipment energy consumption feature extraction using the unsupervised learning center, and calculate the new optimization efficiency of the abnormal data of the refrigeration unit module using the optimization efficiency, initial information and secondary information; The optimization efficiency module is also used to calculate the optimization efficiency of the unsupervised learning center of the data control platform of each refrigeration unit module for real-time equipment energy consumption feature extraction, and to use the optimization efficiency of the unsupervised learning center of the data control platform of each refrigeration unit module for real-time equipment energy consumption feature extraction to calculate the optimization efficiency of the unsupervised learning center for real-time equipment energy consumption feature extraction, wherein the optimization efficiency of the unsupervised learning center for real-time equipment energy consumption feature extraction is the optimization efficiency of the unsupervised learning center of the data control platform of each refrigeration unit module for real-time equipment energy consumption feature extraction and the algorithm of the optimization efficiency difference.

6. A system utilizing the method for optimizing energy consumption of a chiller unit based on an unsupervised learning algorithm according to claim 5, characterized in that: The optimization efficiency management module is used to notify the refrigeration unit operation status information construction module, the unsupervised learning center real-time equipment energy consumption feature extraction module and the optimization efficiency module when the optimization efficiency calculated by the unsupervised learning center for real-time equipment energy consumption feature extraction is not greater than a preset energy consumption threshold, so as to recalculate the operation and energy consumption parameters of the equipment in the refrigeration unit, use the unsupervised learning center for real-time equipment energy consumption feature extraction and calculate the optimization efficiency.

7. A system utilizing the method for optimizing energy consumption of a chiller unit based on an unsupervised learning algorithm according to claim 5, characterized in that: It further includes an energy consumption threshold module for feature extraction of the energy consumption threshold used by the optimization efficiency management module for optimization efficiency comparison.

8. A system utilizing the method for optimizing energy consumption of a chiller unit based on an unsupervised learning algorithm according to claim 7, characterized in that: The refrigeration unit module data transmission module is further used to record the number of times the unsupervised learning center performs real-time equipment energy consumption feature extraction using the command of the optimization efficiency management module, and is used to determine whether the number of times the unsupervised learning center performs real-time equipment energy consumption feature extraction calculation using the module used by the unsupervised learning center to perform real-time equipment energy consumption feature extraction calculation to each refrigeration unit module data control platform has reached a preset number parameter after receiving the command of the optimization efficiency management module that the optimization efficiency of the unsupervised learning center performing real-time equipment energy consumption feature extraction is greater than a preset energy consumption threshold. And when the data transmission module of the refrigeration unit module calculates and the number of times the unsupervised learning center performs real-time equipment energy consumption feature extraction and sends reaches a preset number parameter, it is further used to notify the energy consumption threshold module to perform feature extraction of the energy consumption threshold, wherein the energy consumption threshold module is used to use the optimization efficiency of each unsupervised learning center for real-time equipment energy consumption feature extraction obtained by the optimization efficiency module, and calculate the new energy consumption threshold through the average of the initial energy consumption threshold and the aforementioned optimization efficiency.

Images