CN-116164377-B - Air conditioner running state detection system

CN116164377BCN 116164377 BCN116164377 BCN 116164377BCN-116164377-B

Abstract

The invention discloses an air conditioner running state detection system which comprises an Internet of things system, a data access platform, a service platform, a model archiving server and a service platform, wherein equipment data generated by air conditioner running is reported to the data access platform, the data access platform generates a prediction request and pushes the prediction request to the service platform in a concurrent mode, the service platform accesses the concurrent prediction request of the platform, reads a prediction model from the model archiving server, predicts the air conditioner running state in combination with the equipment data and returns a prediction result to the data access platform, the data access platform reports the prediction result to the service platform, the service platform also receives training tasks of a new model from the service platform, trains the training tasks in an asynchronous multi-process mode and stores the trained prediction model in the model archiving server, and the effects of real-time receiving the data reporting by the equipment, online prediction and online training of the prediction model and updating the model at any time are achieved.

Inventors

WANG YUSHUN

Assignees

青岛海信日立空调系统有限公司

Dates

Publication Date: 20260508
Application Date: 20230315

Claims (7)

1. An air conditioner operation state detection system, characterized by comprising: the internet of things system is used for accessing the air conditioner gateway and receiving equipment data reported by an air conditioner; The data access platform receives the equipment data reported by the air conditioner from the internet system, generates a prediction request based on the equipment data and pushes the prediction request to the service platform; The cache platform is connected with the data access platform and used for caching the matched data; the data access platform identifies and matches the equipment data to obtain the matched data, and generates the prediction request based on the matched data; The service platform receives a training task from the service platform through a training interface and trains the prediction model based on asynchronous multiprocessing, issues a command based on a new model issued by the service platform, issues the trained new prediction model to the cache platform, and judges whether the cache platform has the newly issued prediction model or not before reading the prediction model, if so, reads the newly issued prediction model from the cache platform; The model archiving server is connected with the service platform and used for storing and archiving the prediction data and the trained prediction model; The service platform is used for acquiring a prediction result from the data access platform, and implementing target management by combining the prediction result; The prediction interface comprises: the real-time prediction interface is used for receiving concurrent prediction requests; and the prediction model acquisition interface is used for loading the prediction model from the model archiving server.
2. The system according to claim 1, wherein the service platform reserves a prediction result buffer, performs a plurality of predictions and buffers each prediction result, and returns the prediction result when the plurality of prediction results are identical.
3. The air conditioner operation state detection system according to claim 1, wherein the system further comprises: and the database is used for storing the prediction data and the prediction result, and the service platform reads the prediction result from the database.
4. The system according to claim 1, wherein the data access platform accesses the internet of things system through MQ, MQTT protocol or API interface.
5. The air conditioner operation state detection system according to claim 3, wherein the storing and archiving of the prediction data and the prediction result is implemented using MySQL, postgreSQL, mongoDB or elastic search.
6. The air conditioner operation state detection system according to claim 1, wherein pressure, gas and/or temperature data in the equipment data is recalled CoolProp library implementation.
7. The system according to claim 1, wherein the service platform employs a ASGI framework of Phthon to implement concurrent access of predictive requests and a celery framework of Phthon to implement asynchronous multi-process training tasks.

Description

Air conditioner running state detection system Technical Field The invention relates to the technical field of central air conditioners, in particular to an air conditioner running state detection system. Background After-sales service after air conditioner installation is basically carried out based on a mode of repairing air conditioner faults by clients, and is carried out based on a mode of passive execution of an after-sales service department by active starting of clients. However, the customers are not professional enough and can only feed back the fault phenomenon of the operation of the air conditioner, so that after-sales personnel cannot accurately position the fault position and type, after-sales personnel are required to check to determine the fault position and type, if the faults cannot be removed on site, the after-sales personnel are required to enter the home for maintenance, the after-sales service steps are complicated and the efficiency is low, the use experience of the customers is reduced, and the after-sales service instructions are also reduced. In order to predict the faults of the air conditioner in advance and locate the fault types and positions, a plurality of air conditioner fault repair methods are proposed in the prior art, collect the running data of the air conditioner, predict and diagnose the running state of the air conditioner based on a prediction model or a diagnosis algorithm, on one hand, discover and actively initiate a maintenance program in time at the early stage of the faults, and on the other hand, locate the fault types and fault positions which occur or are about to occur, so that the faults of the air conditioner can be prevented from being discovered only when the faults are accumulated to be unusable, the user experience is reduced, and the method is beneficial to maintenance staff to purposefully and fully prepare to maintain the faults. However, this failure report method needs to collect a large amount of operation data and needs support of multiple prediction or diagnosis models, which all put higher demands on the system of failure prediction diagnosis. The above information disclosed in this background section is only for enhancement of understanding of the background section of the application and therefore it may not form the prior art that is already known to those of ordinary skill in the art. Disclosure of Invention Aiming at the problems pointed out in the background art, the invention provides an air conditioner running state detection system, which constructs a detection system consisting of an Internet of things system, a data access platform, a service platform, a model archiving server and a service platform, gives perfect system support to the current fault diagnosis algorithm, and realizes the effects of reporting data by a real-time receiving device, online prediction, online optimization of a prediction model and issuing a new model at any time. In order to achieve the aim of the invention, the invention is realized by adopting the following technical scheme: an air conditioner operation state detection system is proposed, comprising: the internet of things system is used for accessing the air conditioner gateway and receiving equipment data reported by an air conditioner; The data access platform receives the equipment data reported by the air conditioner from the internet system, generates a prediction request based on the equipment data and pushes the prediction request to the service platform; the service platform receives concurrent prediction requests from the data access platform through a prediction interface, reads the prediction model to predict the running state of the air conditioner, and returns a prediction result to the data access platform; The model archiving server is connected with the service platform and used for storing and archiving the prediction data and the trained prediction model; the service platform is used for acquiring a prediction result from the data access platform, implementing target management in combination with the prediction result, and issuing a prediction model training instruction to the service platform. Compared with the prior art, the air conditioner running state detection system has the advantages that a detection system consisting of an Internet of things system, a data access platform, a service platform, a model archiving server and a service platform is built, equipment data generated by air conditioner running is reported to the data access platform based on the Internet of things system, the data access platform generates a prediction request according to the equipment data and pushes the prediction request to the service platform in a concurrent mode, the service platform receives the concurrent prediction request of the data access platform through a prediction interface, reads a prediction model from the model archiving server, predicts the air conditioner running state by com