Cold machine control method and device, cold machine equipment and heating and ventilation system

Abstract

The invention relates to a refrigerator control method and device, refrigerator equipment and a heating and ventilation system.The refrigerator control method comprises the steps that a plurality of alternative models are determined; determining the score of the alternative model based on the current operation scene and a plurality of score dimensions; the scoring dimensions at least comprise a precision scoring dimension, a speed scoring dimension and a computing power scoring dimension; and a target model is selected from the alternative models based on the score, and operation of the refrigerator is controlled based on the target model. Thus, in the intelligent control process of the refrigerator, the alternative models are evaluated through the precision, speed and computing power dimensions, the model most suitable for the current operation scene can be more accurately selected to intelligently control operation of the refrigerator, and the model selection flexibility in different scenes is improved.

Description

Technical Field

[0001] The present application relates to the field of intelligent control technology, and in particular to a refrigeration machine control method, device, refrigeration machine equipment and HVAC system. Background Art

[0002] In related technologies, chillers, as key equipment in HVAC systems, introduce intelligent control based on artificial intelligence (AI) at the control level. However, in actual applications, different scenarios for chiller operation (such as high-precision control scenarios, low computing resource scenarios, etc.) have different requirements for models, and the flexibility and adaptability of model application are relatively poor. Summary of the invention

[0003] The present application provides a refrigeration machine control method, device, refrigeration machine equipment and HVAC system to solve the technical problem of poor model application flexibility.

[0004] In a first aspect, the present application provides a refrigeration machine control method, comprising: determining multiple alternative models; determining the scores of the alternative models based on the current operating scenario and multiple scoring dimensions; the scoring dimensions include at least: accuracy scoring dimension, speed scoring dimension and computing power scoring dimension; selecting a target model from the multiple alternative models based on the scores, and controlling the operation of the refrigeration machine based on the target model.

[0005] In one possible implementation, determining the score of the alternative model based on the current operating scenario and multiple scoring dimensions includes: determining a sub-score of the alternative model under each scoring dimension based on the current operating scenario and at least one scoring indicator corresponding to each scoring dimension; normalizing the sub-scores; and determining the score of the alternative model based on the normalized sub-scores and weights corresponding to the scoring dimensions.

[0006] In one possible implementation, determining the sub-score of the alternative model under each scoring dimension based on the current operating scenario and at least one scoring indicator corresponding to each scoring dimension includes: performing a predetermined operation through the alternative model; determining the sub-score of the alternative model under each scoring dimension based on the current operating scenario, the operation result of the predetermined operation, and at least one scoring indicator corresponding to each scoring dimension.

[0007] In a possible implementation, the sub-score of the candidate model under each scoring dimension is determined based on the current operating scenario, the operation result of the predetermined operation, and at least one scoring indicator corresponding to each scoring dimension, including: determining the first sub-score of the candidate model under the accuracy scoring dimension based on the current operating scenario, the amount of data corresponding to the predetermined operation, the actual value and the model prediction value corresponding to the operation result of the predetermined operation, and the first scoring indicator corresponding to the accuracy scoring dimension; determining the second sub-score of the candidate model under the speed scoring dimension based on the current operating scenario, the training start and end times and the prediction start and end times corresponding to the operation result of the predetermined operation, and the second scoring indicator corresponding to the speed scoring dimension; determining the third sub-score of the candidate model under the computing power scoring dimension based on the current operating scenario, the system performance parameters corresponding to the operation result of the predetermined operation, and the third scoring indicator corresponding to the computing power scoring dimension.

[0008] In one possible embodiment, after controlling the operation of the refrigeration machine based on the target model, the method further includes: determining whether the target model needs to be updated; in response to the need to update the target model, determining whether to maintain the current working state based on the current working state of the refrigeration machine; if it is determined not to maintain the current working state, controlling the operation of the refrigeration machine based on a predetermined control algorithm and suspending the control function of the target model; and updating the target model.

[0009] In one possible implementation, updating the target model includes: updating and deploying the target model based on the updated model; determining whether the update deployment is successful; if the update deployment fails, restoring the control function of the target model and outputting a prompt message; if the update deployment is successful, controlling the operation of the cold machine based on the updated target model.

[0010] In one possible implementation, determining a plurality of alternative models includes: processing predetermined training data based on a predetermined mechanism model; inputting a first result output by the predetermined mechanism model and the predetermined training data into a predetermined data-driven model; and performing training based on the first result and a second result output by the predetermined data-driven model to obtain a plurality of alternative models for refrigeration machine control.

[0011] In a possible implementation, the processing of predetermined training data based on a predetermined mechanism model includes: inputting the predetermined training data into a predetermined mechanism model to obtain a plurality of refrigeration machine control processing data; and outputting a first result characterizing refrigeration machine performance based on the plurality of refrigeration machine control processing data.

[0012] In a second aspect, the present application provides a refrigeration machine control device, which includes: a determination unit for determining multiple alternative models; a scoring unit for determining the scores of the alternative models based on the current operating scenario and multiple scoring dimensions; the scoring dimensions include at least: an accuracy scoring dimension, a speed scoring dimension, and a computing power scoring dimension; an execution unit for selecting a target model from the multiple alternative models based on the scores, and controlling the operation of the refrigeration machine based on the target model.

[0013] In a third aspect, the present application provides a refrigeration machine device, comprising the refrigeration machine control device described in any one of the aforementioned second aspects.

[0014] In a fourth aspect, the present application provides a HVAC system, comprising the refrigeration equipment described in any one of the aforementioned third aspects.

[0015] In a fifth aspect, the present application provides an electronic device, a processor, a communication interface, a memory and a communication bus, wherein the processor, the communication interface and the memory communicate with each other through the communication bus; the memory is used to store a computer program; and the processor is used to implement any one of the methods described in the first aspect when executing the computer program.

[0016] In a sixth aspect, the present application further provides a storage medium on which a computer program is stored, and when the computer program is executed by a processor, the method described in any one of the first aspects above is implemented.

[0017] The above technical solution provided by the embodiment of the present application has the following advantages over the prior art: the method provided by the embodiment of the present application determines multiple candidate models; determines the score of the candidate model based on the current operating scenario and multiple scoring dimensions; the scoring dimensions include at least: accuracy scoring dimension, speed scoring dimension and computing power scoring dimension; selects a target model from the multiple candidate models based on the score, and controls the operation of the chiller based on the target model. In this way, according to the current operating scenario and the scoring mechanism of multiple dimensions of accuracy, speed, and computing power, the control model that best suits the current operating scenario requirements can be selected, so that when the chiller operation control is performed based on the control model, the operating status of the chiller can be flexibly adapted to improve the accuracy of control and operating efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present application and, together with the description, serve to explain the principles of the present application.

[0019] In order to more clearly illustrate the embodiments of the present application 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, for ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.

[0020] One or more embodiments are exemplarily described by pictures in the corresponding drawings, and these exemplified descriptions do not constitute limitations on the embodiments. Elements with the same reference numerals in the drawings represent similar elements, and unless otherwise stated, the figures in the drawings do not constitute proportional limitations.

[0021] Figure 1 A schematic flow chart of a chiller control method provided in an embodiment of the present application;

[0022] Figure 2 A schematic flow chart of a chiller control method provided in an embodiment of the present application;

[0023] Figure 3 A schematic flow chart of a chiller control method provided in an embodiment of the present application;

[0024] Figure 4 A schematic flow chart of a chiller control method provided in an embodiment of the present application;

[0025] Figure 5 A schematic diagram of a cold machine model training process provided in an embodiment of the present application;

[0026] Figure 6 A schematic diagram of a process for scoring and optimizing a refrigeration machine model provided in an embodiment of the present application;

[0027] Figure 7 A schematic diagram of a process for updating and deploying a cold machine model provided in an embodiment of the present application;

[0028] Figure 8 A schematic diagram of a chiller control device provided in an embodiment of the present application;

[0029] Fig. 9 A schematic diagram of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0030] In order to make the purpose, technical scheme and advantages of the embodiments of the present application clearer, the technical scheme in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, rather than all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in the field without making creative work are within the scope of protection of the present application. It should be noted that the embodiments in the present application and the features in the embodiments can be combined with each other without conflict.

[0031] The disclosure below provides many different embodiments or examples to realize the different structures of the present application. In order to simplify the disclosure of the present application, the parts and settings of specific examples are described below. Of course, they are only examples, and the purpose is not to limit the present application. In addition, the present application can repeat reference numbers and / or letters in different examples. This repetition is for the purpose of simplification and clarity, and does not itself indicate the relationship between the various embodiments and / or settings discussed.

[0032] In order to solve the technical problem of poor flexibility in model application in the prior art, the present application provides a refrigeration machine control method, which scores alternative models according to the current operating scenario and multiple scoring dimensions, so as to select the model that best suits the current scenario to control the refrigeration machine operation, thereby improving the flexibility and accuracy of model application.

[0033] Figure 1 A flow chart of a chiller control method provided in this embodiment is shown as follows: Figure 1 As shown, the method may include:

[0034] S10: Determine multiple candidate models;

[0035] S20: Determine the score of the candidate model based on the current operation scenario and multiple scoring dimensions; the scoring dimensions include at least: accuracy scoring dimension, speed scoring dimension and computing power scoring dimension;

[0036] S30: Selecting a target model from a plurality of candidate models based on the scores, and controlling the operation of the chiller based on the target model.

[0037] The chiller control method provided in this embodiment can be applied to a chiller, or can also be applied to a server or other terminal that has control capabilities for the chiller, such as a terminal device that is connected to the chiller via Bluetooth, a local area network, or other means to control the operation of the chiller. Here, the chiller can be a device used in a HVAC system to achieve functions such as cooling and temperature reduction.

[0038] In this embodiment, determining a plurality of candidate models may include: determining a plurality of candidate models in response to the cold machine being awakened or the cold machine starting to run, wherein being awakened may refer to switching from a sleep state or a shutdown state to a working state or a startup state, etc.

[0039] In one embodiment, the candidate model and the target model may also be referred to as a chiller model or a chiller control model, etc. The candidate model may be a model for chiller control, for example, different candidate models may have different parameters or may contain different sub-model weight distributions. Here, the sub-models of the candidate model may include a predetermined mechanism model and a predetermined data-driven model.

[0040] For example, the predetermined mechanism model can be used to provide constraints on the physical characteristics of the system, and the predetermined data-driven model can be used to supplement the nonlinear characteristics in high-dimensional complex environments. Under different operating conditions of the chiller, the weights of the two can be adjusted dynamically. For example, when the equipment is close to the rated operating condition, the weight of the predetermined mechanism model is higher; and under complex fluctuating conditions, the weight of the predetermined data-driven model is higher.

[0041] In one embodiment, determining the score of the candidate model based on the current operating scenario and multiple scoring dimensions may include: determining a sub-score of the candidate model under each scoring dimension based on the current operating scenario; and determining the score of the candidate model based on the sub-score corresponding to each scoring dimension. The current operating scenario may include at least one of the current operating mode, current operating state, and current operating condition of the refrigerator.

[0042] In one embodiment, determining the sub-score of the alternative model under each scoring dimension based on the current operating scenario may include: performing a predetermined operation through the alternative model, and determining the sub-score of the alternative model under each scoring dimension based on the current operating scenario, the operation result of the predetermined operation, and at least one scoring indicator corresponding to each scoring dimension.

[0043] In one embodiment, determining the score of the candidate model based on the sub-score corresponding to each scoring dimension may include: determining the score of the candidate model based on the sub-score corresponding to each scoring dimension and the weight corresponding to each scoring dimension. For example, it may include: normalizing the sub-scores; and determining the score of the candidate model based on the normalized sub-scores and the weights corresponding to the scoring dimensions.

[0044] In one embodiment, the scoring dimension includes at least one of the following: accuracy scoring dimension, speed scoring dimension, and computing power scoring dimension. The accuracy scoring dimension can be used to evaluate the prediction accuracy of the candidate model, the speed scoring dimension can be used to evaluate the training speed and / or prediction speed of the candidate model, and the computing power scoring dimension can be used to evaluate the occupation of computing power resources by the candidate model during operation.

[0045] In one embodiment, each scoring dimension may correspond to at least one scoring indicator. For example, the first scoring indicator corresponding to the accuracy scoring dimension may include at least one of the following: data variance understanding coefficient, maximum deviation, average deviation, and relative error, etc.; the second scoring indicator corresponding to the speed scoring dimension may include at least one of the following: training time and prediction time, etc.; the third scoring indicator corresponding to the computing power scoring dimension may include at least one of the following: floating point calculation amount, processor utilization, and memory occupancy, etc.

[0046] In one embodiment, selecting a target model from the plurality of candidate models based on the scores may include: selecting the candidate model with the highest score from the plurality of candidate models as the target model.

[0047] In one embodiment, selecting a target model from the plurality of candidate models based on the score may include: selecting a target model from the plurality of candidate models based on the current operating scenario and the score. For example, it may include: when the number of the candidate models with the highest scores among the plurality of candidate models is greater than 1, selecting the candidate model with the highest matching degree with the current operating scenario among the candidate models with the highest scores as the target model.

[0048] Here, the highest degree of match with the current operating scenario may refer to the highest degree of match between the weight distribution of the included sub-models and the current operating scenario.

[0049] In one embodiment, controlling the operation of the refrigerator based on the target model may include: executing a control function based on the target model to control the operation of the refrigerator.

[0050] In one embodiment, controlling the operation of the refrigeration machine based on the target model may include: determining whether there is a current model for controlling the operation of the refrigeration machine; if not, controlling the operation of the refrigeration machine based on the target model; if so, updating the current model based on the target model.

[0051] In one embodiment, updating the current model based on the target model may include: determining whether to maintain the current working state based on the current working state of the refrigerator; if it is determined not to maintain the current working state, controlling the operation of the refrigerator based on a predetermined control algorithm, suspending the control function of the current model, and updating the current model based on the target model.

[0052] In one embodiment, updating the current model based on the target model in the above steps may include: updating and deploying the current model based on the target model; determining whether the update deployment is successful; if the update deployment fails, restoring the control function of the current model and outputting a prompt message; if the update deployment is successful, controlling the operation of the chiller based on the target model.

[0053] In one embodiment, determining whether to maintain the current working state may include determining whether the refrigerator is allowed to maintain the current working state. Maintaining the current working state may refer to maintaining the current operating conditions, with all operating parameters remaining unchanged without any adjustment.

[0054] In one embodiment, the method may further include: if it is determined to maintain the current working state, keeping the current operating parameters unchanged, and updating the current model based on the target model.

[0055] In one embodiment, determining whether to maintain the current working state based on the current working state of the refrigeration machine may include: determining whether the current operating parameters need to be changed based on the current working state of the refrigeration machine; if they need to be changed, determining not to maintain the current working state; if they do not need to be changed, determining to maintain the current working state.

[0056] Here, determining whether the current operating parameters need to be changed may include: determining whether the current operating parameters need to be changed within a predetermined time period. The predetermined time period may be related to the time period required to update the current model based on the target model.

[0057] In one embodiment, controlling the operation of the refrigerator based on a predetermined control algorithm may include: switching to a predetermined group control strategy, and taking over the operation control of the refrigerator through a predetermined control algorithm. The predetermined control algorithm here may be a conventional control algorithm.

[0058] In this way, based on the current operating scenario and the scoring mechanism in multiple dimensions of accuracy, speed, and computing power, the control model that best matches the current operating scenario requirements can be selected. Therefore, when the refrigeration machine operation control is performed based on this control model, the operating conditions of the refrigeration machine can be flexibly adapted to improve the control accuracy and operating efficiency.

[0059] In order to comprehensively evaluate these scores, make the scores of all dimensions comparable within the same dimension and balance the different importance of different dimensions, therefore, Figure 2 A flow chart of a chiller control method provided in this embodiment is shown as follows: Figure 2 As shown, in the above step S20, determining the score of the candidate model based on the current operating scenario and multiple scoring dimensions may include:

[0060] S21: Determine a sub-score of the candidate model under each scoring dimension based on the current operating scenario and at least one scoring indicator corresponding to each scoring dimension;

[0061] S22: normalize the sub-scores;

[0062] S23: Determine the score of the candidate model based on the normalized sub-scores and the weights corresponding to the scoring dimensions.

[0063] In one embodiment, the sub-scores may correspond to the scoring indicators one by one, that is, each scoring indicator corresponds to one sub-score.

[0064] In one embodiment, normalizing the sub-scores may include: normalizing each sub-score based on a minimum value and a maximum value among all sub-scores.

[0065] For example, the normalization formula is:

[0066] Where: X is the sub-score, X nor is the normalized sub-score, min(X) is the minimum value of all sub-scores, and max(X) is the maximum value of all sub-scores.

[0067] In one embodiment, each scoring dimension corresponds to a weight, for example, the accuracy scoring dimension corresponds to the first weight The speed rating dimension corresponds to the second weight The computing power scoring dimension corresponds to the third weight

[0068] In one embodiment, the first scoring indicator corresponding to the accuracy scoring dimension includes: data variance understanding coefficient RS, maximum deviation MD, average deviation AD and relative error AR; the second scoring indicator corresponding to the speed scoring dimension may include: training time TT and prediction time PT; the third scoring indicator corresponding to the computing power scoring dimension may include: floating-point calculation amount FL, processor utilization GC and memory occupancy MU.

[0069] Correspondingly, the sub-scores corresponding to the first scoring indicator under the accuracy scoring dimension after normalization are: RS nor ,MD nor , A.D. nor , AR nor ; After normalization, the sub-scores corresponding to the second scoring indicator under the speed scoring dimension are: TT nor , PT nor ; After normalization, the sub-scores corresponding to the third scoring indicator under the computing power scoring dimension are: FL nor , GC nor MU nor .

[0070] In one embodiment, step S23 may include: determining a first product of the sum of the sub-scores corresponding to all first scoring indicators under the accuracy scoring dimension and the first weight; determining a second product of the sum of the sub-scores corresponding to all second scoring indicators under the speed scoring dimension and the second weight; determining a third product of the sum of the sub-scores corresponding to all third scoring indicators under the computing power scoring dimension and the third weight; and calculating the sum of the first product, the second product, and the third product as the score of the alternative model.

[0071] That is, the score of the candidate model

[0072] In this way, the sub-scores of multiple scoring indicators of different dimensions are normalized, and then weighted calculations are performed based on the weights of different dimensions. The resulting scores can be more consistent with the performance of the model in terms of the importance of different dimensions, so that the scores can more accurately describe the model characteristics.

[0073] In some embodiments, step S21 may include: performing a predetermined operation through the alternative model; determining the sub-score of the alternative model under each scoring dimension based on the current operating scenario, the operation result of the predetermined operation, and at least one scoring indicator corresponding to each scoring dimension.

[0074] In one embodiment, performing a predetermined operation through an alternative model may refer to processing predetermined data through the alternative model, such as performing a training operation and / or a prediction operation based on the predetermined data through the alternative model.

[0075] In one embodiment, the operation result of the predetermined operation may refer to the actual value and the model prediction value corresponding to the operation result of the predetermined operation, the training start and end times and the prediction start and end times corresponding to the operation result of the predetermined operation, and at least one of the system performance parameters corresponding to the operation result of the predetermined operation.

[0076] Here, the actual value is the actual value corresponding to the predetermined data, and the model prediction value is the predicted value output by the model when a prediction operation is performed based on the predetermined data.

[0077] In this way, the actual operating performance of the model is reflected through the model's processing results, and then combined with multiple indicators in different dimensions, the scoring can more comprehensively and three-dimensionally characterize the characteristics of the model itself.

[0078] In one embodiment, determining the sub-score of the candidate model under each scoring dimension based on the current operation scenario, the operation result of the predetermined operation, and at least one scoring indicator corresponding to each scoring dimension includes:

[0079] Step a: determining a first sub-score of the candidate model under the accuracy scoring dimension based on the current running scenario, the amount of data corresponding to the predetermined operation, the actual value and the model prediction value corresponding to the operation result of the predetermined operation, and the first scoring indicator corresponding to the accuracy scoring dimension;

[0080] Step b: determining a second sub-score of the candidate model under the speed scoring dimension based on the current running scenario, the training start and end times and the predicted start and end times corresponding to the operation result of the predetermined operation, and the second scoring indicator corresponding to the speed scoring dimension;

[0081] Step c: Based on the current operating scenario, the system performance parameters corresponding to the operation result of the predetermined operation, and the third scoring indicator corresponding to the computing power scoring dimension, determine the third sub-score of the alternative model under the computing power scoring dimension.

[0082] In one embodiment, each first scoring indicator corresponds to a first sub-score, and the corresponding first scoring indicators under the accuracy scoring dimension include: data variance understanding coefficient RS, maximum deviation MD, average deviation AD and relative error AR.

[0083] Accordingly, step a may include: based on the current running scenario, the amount of data n corresponding to the predetermined operation, and the actual value y corresponding to the operation result of the predetermined operation i With the model prediction value and the average of the actual values Determine the first sub-score corresponding to the data variance understanding coefficient RS:

[0084]

[0085] Here, the data variance understanding coefficient indicates the alternative model's ability to understand the data variance. The closer the coefficient is to 1, the better the understanding ability.

[0086] Correspondingly, step a may also include: based on the current running scenario, the actual value y corresponding to the operation result of the predetermined operation i With the model prediction value The maximum difference between the two, the first sub-score corresponding to the maximum deviation MD is determined as: Here, the maximum deviation represents the error of the alternative model under the most unfavorable conditions.

[0087] Accordingly, step a may also include: based on the current running scenario, the amount of data n corresponding to the predetermined operation, and the actual value y corresponding to the operation result of the predetermined operation i With the model prediction value Determine the first sub-score corresponding to the average deviation AD as:

[0088]

[0089] Accordingly, step a may also include: based on the current running scenario, the amount of data n corresponding to the predetermined operation, and the actual value y corresponding to the operation result of the predetermined operation i With the model prediction value Determine the first sub-score corresponding to the relative error AR as:

[0090]

[0091] In one embodiment, each second scoring indicator corresponds to a second sub-score, and the second scoring indicator corresponding to the speed scoring dimension may include: training time TT and predicted time PT. The training start and end times may include the training start time and the training end time, and the predicted start and end times may include the predicted start time and the predicted end time.

[0092] Accordingly, step b may include: based on the current operation scenario, the training start time TT corresponding to the operation result of the predetermined operation s and training end time TT e , determine the second sub-score corresponding to the training time TT as: TT = TT s -TT e .

[0093] Correspondingly, step b may also include: based on the current operation scenario, the predicted start time PT corresponding to the operation result of the predetermined operation s and predicted end time PT e , determine the second sub-score corresponding to the predicted time PT as: PT = PT s -PT e .

[0094] In one embodiment, each third scoring indicator corresponds to a third sub-score, and the third scoring indicators corresponding to the computing power scoring dimension may include: floating point calculation amount FL, processor utilization GC, and memory occupancy MU.

[0095] Accordingly, the system performance parameter may include the number of floating-point operations. Step c may include: determining a third sub-score corresponding to the floating-point calculation amount FL based on the current running scenario and the number of floating-point operations corresponding to the operation result of the predetermined operation. For example, the number of floating-point operations may be the number of floating-point operations of all operations during the execution of the candidate model, and the FL of the candidate model may be calculated and analyzed across frameworks through the ONNX tool.

[0096] Correspondingly, the system performance parameters may also include: the processor execution operation time, and the total operation time of the alternative model. Step c may also include: based on the current operation scenario, the processor execution operation time corresponding to the operation result of the predetermined operation Tactive And the total running time of the alternative model T total , determine the third sub-score corresponding to the processor utilization GC as:

[0097]

[0098] The processor execution time is the time the processor actually uses to execute operations, which can be obtained through system performance monitoring tools (such as 'nvidia-smi'). The total running time of the candidate model is the total running time of the entire candidate model, including waiting time, data loading time, etc., which can be obtained through log records or system tools.

[0099] Correspondingly, the system performance parameters may also include: the actual amount of memory used and the total amount of system memory. Step c may also include: based on the current running scenario, the actual amount of memory used M corresponding to the operation result of the predetermined operation used And the total system memory M total , determine the third sub-score corresponding to the memory usage MU as:

[0100]

[0101] In this way, calculating scores based on different parameters obtained from the model processing process and results in different dimensions can more accurately characterize the performance characteristics of the model in different dimensions, thereby obtaining more accurate scores that are in line with the operating scenarios.

[0102] After the model is deployed, it needs to be updated or upgraded during use. However, the related technologies often require manual updates and upgrades after shutting down the device to avoid the device from failing to operate during the update process, which is quite troublesome. Figure 3 A flow chart of a chiller control method provided in this embodiment is shown as follows: Figure 3 As shown, after the above step S30 controls the operation of the refrigerator based on the target model, the method may further include:

[0103] S40: Determine whether the target model needs to be updated;

[0104] S50: In response to the need to update the target model, determine whether to maintain the current working state based on the current working state of the refrigerator;

[0105] S60: If it is determined that the current working state is not to be maintained, the operation of the chiller is controlled based on a predetermined control algorithm, and the control function of the target model is suspended;

[0106] S70: Update the target model.

[0107] In one embodiment, determining whether to maintain the current working state may include determining whether the refrigerator is allowed to maintain the current working state. Maintaining the current working state may refer to maintaining the current operating conditions, with all operating parameters remaining unchanged without any adjustment.

[0108] In one embodiment, the method may further include: if it is determined to maintain the current working state, keeping the current operating parameters unchanged, and updating the target model.

[0109] In one embodiment, determining whether to maintain the current working state based on the current working state of the refrigeration machine may include: determining whether the current operating parameters need to be changed based on the current working state of the refrigeration machine; if they need to be changed, determining not to maintain the current working state; if they do not need to be changed, determining to maintain the current working state.

[0110] Here, determining whether the current operating parameters need to be changed may include: determining whether the current operating parameters need to be changed within a predetermined time period. The predetermined time period may be related to the time period required to update the target model.

[0111] In one embodiment, controlling the operation of the refrigerator based on a predetermined control algorithm in step S60 may include: switching to a predetermined group control strategy and taking over the operation control of the refrigerator through a predetermined control algorithm. The predetermined control algorithm here may be a conventional control algorithm.

[0112] In one embodiment, suspending the control function of the target model may include: freezing the control logic of the target model and blocking the adjustment instructions of the target model. Therefore, the system will no longer respond to the adjustment signal brought by the model control, avoiding fluctuations caused by model switching.

[0113] In one embodiment, suspending the control function of the target model may include: suspending the control function of the target model and monitoring the operating state of the chiller during the suspension of the control function of the target model. For example, continuing to monitor the operating state of the equipment to ensure that no abnormal operation or equipment failure occurs during the model update.

[0114] In this way, the control model can be updated and deployed online smoothly during normal operation of the chiller, avoiding the impact of shutdown or interruption of control logic on the overall operation performance of the equipment. Sufficient flexibility is provided so that the model update can be carried out smoothly in different application scenarios, which can greatly reduce the risk of system interruption.

[0115] In one embodiment, updating the target model may include: updating and deploying the target model based on the updated model; determining whether the update deployment is successful; if the update deployment fails, restoring the control function of the target model and outputting a prompt message; if the update deployment is successful, controlling the operation of the cold machine based on the updated target model.

[0116] In one embodiment, determining whether the update deployment is successful may include: determining whether the model update deployment is successful according to the version number of the target model.

[0117] In one embodiment, if the update deployment fails, it may refer to if the update deployment fails and the number of failures reaches a predetermined number of times. For example, the predetermined number of times may be 2 times, 3 times, or 5 times.

[0118] In one embodiment, restoring the control function of the target model and outputting a prompt message may refer to terminating the update deployment and rolling back to the old model, while giving a clear error prompt to notify the user to perform manual intervention.

[0119] In one embodiment, controlling the operation of the chiller based on the updated target model may refer to unfreezing the control logic of the target model and restoring the adjustment instructions of the target model, thereby putting the new model into actual control tasks.

[0120] In this way, if the model update fails, the original model control can be restored in time to avoid affecting the system control for a long time. The model update deployment can be completed smoothly without affecting the current operation. This not only reduces the complexity of system maintenance, but also improves the flexibility and reliability of the chiller AI control, ensuring that the system can operate continuously and efficiently.

[0121] In some embodiments, determining multiple candidate models in step S10 may include:

[0122] Processing predetermined training data based on a predetermined mechanism model;

[0123] Inputting the first result output by the predetermined mechanism model and the predetermined training data into a predetermined data-driven model;

[0124] Training is performed based on the first result and the second result output by the predetermined data-driven model to obtain multiple candidate models for refrigeration machine control.

[0125] In one embodiment, inputting the first result output by the predetermined mechanism model and the predetermined training data into the predetermined data-driven model may include: inputting the first result output by the predetermined mechanism model into the predetermined data-driven model as a constraint condition of the data-driven model; and inputting the predetermined training data into the predetermined data-driven model for training. In this way, the data-driven model can not only rely on historical data, but also follow the physical laws of the equipment, thereby improving the interpretability and robustness of the model.

[0126] In one embodiment, the predetermined data-driven model can be driven by a predetermined algorithm. For example, the predetermined algorithm can be a random forest, a support vector machine (SVM), a K-nearest neighbor, a neural network, a distributed gradient boosting library (XGBoost), or the like.

[0127] In one embodiment, before processing the predetermined training data based on the predetermined mechanism model, the method may further include: importing the predetermined training data; and preprocessing the predetermined training data. Importing the predetermined training data may be obtaining historical operation data corresponding to a predetermined time range as the predetermined training data, or directly uploading the predetermined training data.

[0128] In one embodiment, the preprocessing of the predetermined training data may include removing abnormal values ​​(such as frequent start and stop, abnormal power on and off events, etc.), data denoising and normalization, so as to ensure the quality and consistency of the input data and meet the requirements of subsequent model training.

[0129] In this way, through the hybrid training mechanism of the two models, by using the mechanism model as one of the features of the data-driven model input, the accuracy of model training can be better improved, which is consistent with the equipment operation mechanism and the characteristics of the training data, thereby improving the model performance.

[0130] In one embodiment, the processing of predetermined training data based on a predetermined mechanism model may include:

[0131] Inputting predetermined training data into a predetermined mechanism model to obtain a plurality of refrigeration machine control processing data;

[0132] A first result representing the performance of the chiller is output based on the plurality of chiller control process data.

[0133] In one embodiment, the chiller control processing data may include at least one of the following: a heat exchange efficiency coefficient CAPFT, a heat balance coefficient EIRFT, and a refrigeration load coefficient EIRFPLR.

[0134] In one embodiment, obtaining a plurality of refrigeration machine control processing data may include: determining a plurality of refrigeration machine operation data based on predetermined training data; and determining a plurality of refrigeration machine control processing data based on the refrigeration machine operation data.

[0135] Here, the chiller operation data may include at least one of the following: the main chiller water supply temperature T eo , return water temperature of the main chiller pipe T ei , Cooling main return water temperature T ci , Cooling main return water temperature T co , Refrigeration capacity Q e , Rated cooling capacity of the chiller Q e,ref .

[0136] In one embodiment, determining a plurality of chiller control processing data based on chiller operation data may include: determining a first calculation coefficient corresponding to each chiller operation data based on model fitting;

[0137] A plurality of chiller control process data are determined based on the chiller operation data and the first calculation coefficient.

[0138] For example, CAPFT = α0 + α1T eo +α2T eo 2 +α3T ci +α4T ci 2 +α5T eo T ci ;

[0139] EIRFT=β0+β1T ei +β2T ei 2 +β3T co +β4T co 2 +β5T eo T co ;

[0140] EIRFPLR=γ0+γ1Q e +γ2Q e 2 +γ3T co +γ4T co 2 +γ5Q e T co .

[0141] Among them, α0, α1, α2, α3, α4, α5, β0, β1, β2, β3, β4, β5, γ0, γ1, γ2, γ3, γ4, and γ5 are all first calculation coefficients.

[0142] In one embodiment, outputting a first result representing the performance of the refrigeration machine based on the plurality of refrigeration machine control processing data may include: determining a second calculation coefficient based on model fitting; outputting a result representing the performance of the refrigeration machine based on the plurality of refrigeration machine control processing data and the second calculation coefficient.

[0143] First result.

[0144] For example, the first output result may be the coefficient of performance (COP) of the chiller, expressed as:

[0145]

[0146] Among them, δ COPis the second calculation coefficient.

[0147] In this way, a new performance coefficient expression is introduced during the training process, so that the output

[0148] The results can more accurately and balancedly characterize the performance of different aspects of the refrigerator and improve the reliability of the model.

[0149] As a possible implementation, Figure 4 As shown, a chiller control method is provided.

[0150] Specifically, the method may include:

[0151] 1. Model training module. The model training module includes data processing, model customization selection, parameter optimization and model training, such as Figure 5 As shown, in order to achieve higher prediction accuracy and control effect under different operating conditions, the refrigeration machine model is trained based on the mechanism model and the data-driven model. The specific implementation method is as follows:

[0152] Mechanism model as a constraint condition for data-driven model: When training a data-driven model, the output of the mechanism model can be used as an additional input feature to participate in the training process of the data-driven model. This allows the data-driven model to not only rely on historical data, but also follow the physical laws of the equipment, improving the interpretability and robustness of the model.

[0153] Hybrid model design: A hybrid model framework is formed by training the mechanism model and the data-driven model in parallel. The mechanism model provides constraints on the physical characteristics of the system, while the data-driven model supplements the nonlinear characteristics in high-dimensional complex environments. The weights of the two can be adjusted dynamically in different working ranges of the chiller. For example, when the equipment is close to the rated operating conditions, it relies more on the mechanism model; while under complex fluctuating conditions, it relies more on the predictive ability of the data-driven model.

[0154] Parameter optimization and model training: After the model is selected, the model's hyperparameters are adjusted automatically or semi-automatically (users can customize the parameter range) to optimize model performance and perform training.

[0155] 2. Model optimization module. The model optimization module mainly scores the model from three dimensions: accuracy priority, speed priority, and computing power priority. For example Figure 6 According to the comprehensive scores of each model, the model with the highest comprehensive score is automatically selected as the optimal model and output.

[0156] 3. Model update and deployment module. Ensure that the refrigeration system can smoothly implement online update and deployment of the control model during normal operation to avoid affecting the overall operation performance of the equipment due to shutdown or interruption of control logic. This module includes four main steps, such as Figure 7 shown.

[0157] Through the model deployment module, the refrigeration system can successfully complete the model update deployment without affecting the current operation. This not only reduces the complexity of system maintenance, but also improves the flexibility and reliability of the refrigeration AI control, ensuring that the system can operate continuously and efficiently.

[0158] like Figure 8 As shown, the embodiment of the present application provides a chiller control device, which may include:

[0159] A determination unit 100, configured to determine a plurality of candidate models;

[0160] A scoring unit 200, configured to determine a score of the candidate model based on a current operating scenario and a plurality of scoring dimensions; the scoring dimensions at least include: an accuracy scoring dimension, a speed scoring dimension, and a computing power scoring dimension;

[0161] The execution unit 300 is configured to select a target model from the plurality of candidate models based on the scores, and control the operation of the refrigerator based on the target model.

[0162] In some embodiments, the scoring unit 200 is specifically used to: determine the sub-score of the alternative model under each scoring dimension based on the current operating scenario and at least one scoring indicator corresponding to each scoring dimension; normalize the sub-score; and determine the score of the alternative model based on the normalized sub-score and the weight corresponding to the scoring dimension.

[0163] In one possible implementation, the scoring unit 200 is specifically used to: perform a predetermined operation through the alternative model; and determine a sub-score of the alternative model under each scoring dimension based on the current operating scenario, the operation result of the predetermined operation, and at least one scoring indicator corresponding to each scoring dimension.

[0164] In one possible implementation, the scoring unit 200 is specifically used to: determine the first sub-score of the alternative model under the accuracy scoring dimension based on the current operating scenario, the amount of data corresponding to the predetermined operation, the actual value and the model prediction value corresponding to the operation result of the predetermined operation, and the first scoring indicator corresponding to the accuracy scoring dimension; determine the second sub-score of the alternative model under the speed scoring dimension based on the current operating scenario, the training start and end times and the prediction start and end times corresponding to the operation result of the predetermined operation, and the second scoring indicator corresponding to the speed scoring dimension; determine the third sub-score of the alternative model under the computing power scoring dimension based on the current operating scenario, the system performance parameters corresponding to the operation result of the predetermined operation, and the third scoring indicator corresponding to the computing power scoring dimension.

[0165] In one possible implementation, the execution unit 300 is further used to: determine whether the target model needs to be updated; in response to the need to update the target model, determine whether to maintain the current working state based on the current working state of the refrigerator; if it is determined not to maintain the current working state, control the operation of the refrigerator based on a predetermined control algorithm and suspend the control function of the target model; and update the target model.

[0166] In one possible implementation, the execution unit 300 is specifically used to: update and deploy the target model based on the update model; determine whether the update deployment is successful; if the update deployment fails, restore the control function of the target model and output a prompt message; if the update deployment is successful, control the operation of the cold machine based on the updated target model.

[0167] In one possible implementation, the determination unit 100 is specifically used to: process predetermined training data based on a predetermined mechanism model; input a first result output by the predetermined mechanism model and the predetermined training data into a predetermined data-driven model; and perform training based on the first result and a second result output by the predetermined data-driven model to obtain multiple alternative models for refrigeration machine control.

[0168] In a possible implementation manner, the determination unit 100 is specifically configured to: input predetermined training data into a predetermined mechanism model to obtain a plurality of refrigeration machine control processing data; and output a first result representing refrigeration machine performance based on the plurality of refrigeration machine control processing data.

[0169] An embodiment of the present application further provides a refrigeration device, which includes the refrigeration control device described in any one or more of the aforementioned embodiments.

[0170] An embodiment of the present application further provides a HVAC system, which includes the refrigeration equipment described in any one or more of the aforementioned embodiments.

[0171] like Fig. 9 As shown, an embodiment of the present application provides an electronic device, including a processor 111, a communication interface 112, a memory 113 and a communication bus 114, wherein the processor 111, the communication interface 112, and the memory 113 communicate with each other through the communication bus 114.

[0172] Memory 113, used for storing computer programs;

[0173] In one embodiment of the present application, the processor 111, when being used to execute the program stored in the memory 113, implements the refrigeration machine control method provided by any one or more of the aforementioned method embodiments, which at least includes: determining multiple alternative models; determining the scores of the alternative models based on the current operating scenario and multiple scoring dimensions; the scoring dimensions include at least: accuracy scoring dimension, speed scoring dimension and computing power scoring dimension; selecting a target model from the multiple alternative models based on the scores, and controlling the operation of the refrigeration machine based on the target model.

[0174] An embodiment of the present application further provides a computer-readable storage medium having a computer program stored thereon, and when the computer program is executed by a processor, the steps of the cold machine control method provided by any one or more of the aforementioned method embodiments are implemented.

[0175] The device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0176] Through the description of the above implementation methods, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a general hardware platform, and of course, by hardware. Based on this understanding, the above technical solution is essentially or the part that contributes to the relevant technology can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0177] It should be understood that the terms used in the text are only for the purpose of describing specific example embodiments, and are not intended to be limiting. Unless the context clearly indicates otherwise, the singular forms "one", "an" and "said" as used in the text may also be meant to include plural forms. The terms "include", "comprise", "contain", and "have" are inclusive, and therefore specify the existence of stated features, steps, operations, elements and / or parts, but do not exclude the existence or addition of one or more other features, steps, operations, elements, parts, and / or combinations thereof. The method steps, processes, and operations described herein are not interpreted as necessarily requiring them to be performed in the specific order described or illustrated, unless the execution order is clearly indicated. It should also be understood that additional or alternative steps may be used.

[0178] In the absence of contradiction, each step in a certain implementation mode or example can be implemented as an independent example, and the steps can be arbitrarily combined. For example, a solution after removing some steps in a certain implementation mode or example can also be implemented as an independent example, and the order of the steps in a certain implementation mode or example can be arbitrarily exchanged. In addition, the optional methods or optional examples in a certain implementation mode or example can be arbitrarily combined; in addition, the various implementation modes or examples can be arbitrarily combined. For example, some or all steps of different implementation modes or examples can be arbitrarily combined, and a certain implementation mode or example can be arbitrarily combined with the optional methods or optional examples of other implementation modes or examples.

[0179] The above description is only a specific implementation of the present application, so that those skilled in the art can understand or implement the present application. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein can be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application will not be limited to the embodiments shown herein, but will conform to the widest range consistent with the principles and novel features applied for herein.

Claims

1. A chiller control method, characterized in that: The method comprises: Identify multiple candidate models; Determine the score of the candidate model based on the current operation scenario and multiple scoring dimensions; the scoring dimensions include at least: accuracy scoring dimension, speed scoring dimension and computing power scoring dimension; A target model is selected from the plurality of candidate models based on the scores, and the operation of the chiller is controlled based on the target model.

2. The method according to claim 1, characterized in that The step of determining the score of the candidate model based on the current operation scenario and multiple scoring dimensions includes: Determine a sub-score of the candidate model under each scoring dimension based on the current operating scenario and at least one scoring indicator corresponding to each scoring dimension; normalizing the sub-scores; The score of the candidate model is determined based on the normalized sub-scores and the weights corresponding to the scoring dimensions.

3. The method according to claim 2, characterized in that The determining, based on the current operating scenario and at least one scoring indicator corresponding to each scoring dimension, a sub-score of the candidate model under each scoring dimension includes: Execute a predetermined operation through the alternative model; Based on the current operating scenario, the operation result of the predetermined operation and at least one scoring indicator corresponding to each scoring dimension, a sub-score of the candidate model under each scoring dimension is determined.

4. The method according to claim 3, characterized in that The determining, based on the current operation scenario, the operation result of the predetermined operation, and at least one scoring indicator corresponding to each scoring dimension, the sub-score of the candidate model under each scoring dimension includes: Determine a first sub-score of the candidate model under the accuracy scoring dimension based on the current operating scenario, the amount of data corresponding to the predetermined operation, the actual value and the model prediction value corresponding to the operation result of the predetermined operation, and the first scoring indicator corresponding to the accuracy scoring dimension; Determine a second sub-score of the candidate model under the speed scoring dimension based on the current running scenario, the training start and end times and the predicted start and end times corresponding to the operation result of the predetermined operation, and the second scoring indicator corresponding to the speed scoring dimension; Based on the current operating scenario, the system performance parameters corresponding to the operation result of the predetermined operation, and the third scoring indicator corresponding to the computing power scoring dimension, the third sub-score of the alternative model under the computing power scoring dimension is determined.

5. The method according to claim 1, characterized in that After controlling the operation of the chiller based on the target model, the method further includes: Determining whether the target model needs to be updated; In response to the need to update the target model, determining whether to maintain the current operating state based on the current operating state of the refrigerator; If it is determined that the current working state is not to be maintained, the operation of the refrigerator is controlled based on a predetermined control algorithm, and the control function of the target model is suspended; The target model is updated.

6. The method according to claim 5, characterized in that The updating of the target model comprises: Update and deploy the target model based on the update model; Determine if the update deployment was successful; If the update deployment fails, the control function of the target model is restored and a prompt message is output; If the update deployment is successful, the operation of the refrigerator is controlled based on the updated target model.

7. The method according to claim 1, characterized in that The determining of multiple candidate models comprises: Processing predetermined training data based on a predetermined mechanism model; Inputting the first result output by the predetermined mechanism model and the predetermined training data into a predetermined data-driven model; Training is performed based on the first result and the second result output by the predetermined data-driven model to obtain multiple candidate models for refrigeration machine control.

8. The method according to claim 7, characterized in that The processing of predetermined training data based on a predetermined mechanism model includes: Inputting predetermined training data into a predetermined mechanism model to obtain a plurality of refrigeration machine control processing data; A first result representing the performance of the chiller is output based on the plurality of chiller control process data.

9. A cooling machine control device, characterized in that: The device comprises: A determination unit, used for determining a plurality of candidate models; A scoring unit, configured to determine a score of the candidate model based on a current operating scenario and a plurality of scoring dimensions; the scoring dimensions at least include: an accuracy scoring dimension, a speed scoring dimension, and a computing power scoring dimension; An execution unit is used to select a target model from the plurality of candidate models based on the scores, and control the operation of the refrigerator based on the target model.

10. A cooling machine device, characterized in that: The refrigeration equipment includes the refrigeration control device according to claim 9.

11. A HVAC system, characterized in that: The HVAC system includes the refrigeration equipment according to claim 10.

12. An electronic device, characterized in that: include: A processor, a communication interface, a memory and a communication bus, wherein the processor, the communication interface and the memory communicate with each other via the communication bus; The memory is used to store a computer program; and the processor is used to implement the method according to any one of claims 1 to 8 when executing the computer program.

13. A storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the method according to any one of claims 1 to 8 is implemented.

Images