CN-121977284-A - Big data-based air purifier driving control method, device, equipment, medium and program product

CN121977284ACN 121977284 ACN121977284 ACN 121977284ACN-121977284-A

Abstract

The application relates to an air purifier driving control method, a device, a computer device, a storage medium and a computer program product based on big data. The method comprises the steps of obtaining real-time sensing data of target equipment and running state data of associated equipment, obtaining external environment time sequence data through a preset data interface, preprocessing, extracting time sequence feature vectors and context feature vectors related to dynamic change of air quality, processing the time sequence feature vectors and the context feature vectors based on a pre-built air quality prediction model, obtaining a concentration prediction sequence of target pollutants in a future preset time window, and solving through a preset optimization algorithm by taking the concentration prediction sequence as input and taking a target concentration threshold value at a position and minimum energy consumption as constraint targets to obtain a control instruction sequence of the target equipment, wherein the control instruction sequence is based on the control instruction sequence to drive the target equipment. The method can enhance the energy efficiency and the purification efficiency of the air purifier.

Inventors

ZHANG YI
WANG XIAOMING
WANG LICHONG
LI ZHIYONG

Assignees

北京三五二环保科技有限公司

Dates

Publication Date: 20260505
Application Date: 20260202

Claims (10)

1. An air purifier driving control method based on big data, which is characterized in that the method comprises the following steps: Acquiring real-time sensing data of target equipment and running state data of associated equipment, and acquiring external environment time sequence data through a preset data interface; preprocessing the real-time sensing data, the real-time running state data and the external environment time sequence data, extracting time sequence feature vectors and context feature vectors related to air quality dynamic change, wherein the preprocessing comprises data cleaning, data alignment and normalization processing; processing the time sequence feature vector and the context feature vector based on a pre-constructed air quality prediction model to obtain a concentration prediction sequence of the target pollutant in a future preset time window; and solving by a preset optimization algorithm by taking the concentration prediction sequence as input and taking a target concentration threshold value at a position and minimized energy consumption as constraint targets to obtain a control instruction sequence of the target equipment, and driving the target equipment based on the control instruction sequence.
2. The method according to claim 1, wherein the obtaining the control instruction sequence of the target device by solving through a preset optimization algorithm with the concentration prediction sequence as input and the target concentration threshold at the bit and the minimum energy consumption as constraint targets includes: Constructing a multi-objective cost function comprising a fan power consumption function, a noise radiation function and a pollutant concentration deviation penalty function; And (3) adopting a model predictive control framework, carrying out online optimization on the multi-objective cost function in a rolling time domain, and solving to obtain a fan rotating speed control sequence of the target equipment in a future preset time domain as the control instruction sequence.
3. The method according to claim 2, wherein the obtaining the control instruction sequence of the target device by solving through a preset optimization algorithm with the concentration prediction sequence as input and the target concentration threshold at the bit and the minimum energy consumption as constraint targets further comprises: training a personalized preference model based on the user history interaction data to generate a user-specific concentration comfort interval and a noise sensitivity coefficient; and substituting the concentration comfort interval and the noise sensitivity coefficient as dynamic parameters into the multi-objective cost function to solve.
4. The method of claim 1, wherein preprocessing the real-time sensing data, real-time operating state data, and the external environment time series data and extracting time series feature vectors and context feature vectors related to dynamic changes in air quality comprises: Aiming at the sensing data flow, a Z-Score algorithm or a quartile range method based on a sliding window is adopted to identify and reject numerical outliers; Unifying data from different sources to a preset standard time stamp reference, and mapping external environment time sequence data to specific indoor space coordinates where the target equipment is located through a preset geographic grid interpolation algorithm; and processing the numerical type features with different dimensions by adopting a maximum and minimum value normalization or Z-Score normalization method to generate a normalized feature vector matched with the input of the air quality prediction model.
5. The method of claim 1, wherein the processing the timing feature vector and the context feature vector based on the pre-constructed air quality prediction model, after obtaining a predicted sequence of concentration of the target contaminant within a future pre-set time window, further comprises: Taking new multi-source heterogeneous data and a corresponding pollutant concentration true value as an incremental training sample based on a preset period, and performing online fine adjustment on the air quality prediction model; and continuously monitoring the prediction error of the air quality prediction model, and when the prediction error continuously exceeds a set threshold value, judging that the performance of the air quality prediction model drifts, and triggering the air quality prediction model to be retrained and deployed by using a historical complete data set.
6. The method of claim 1, wherein the pre-trained air quality prediction model is a sequence prediction model constructed based on a long-term memory network or a time-series convolution network, and sample data is historical multi-source heterogeneous data and a corresponding historical pollutant concentration truth sequence.
7. An air cleaner drive control device based on big data, the device comprising: the data acquisition module is used for acquiring real-time sensing data of the target equipment and running state data of the associated equipment, and acquiring external environment time sequence data through a preset data interface; The data processing module is used for preprocessing the real-time sensing data, the real-time running state data and the external environment time sequence data, extracting time sequence feature vectors and context feature vectors related to the dynamic change of the air quality, wherein the preprocessing comprises data cleaning, data alignment and normalization processing; The air quality prediction module is used for processing the time sequence feature vector and the context feature vector based on a pre-constructed air quality prediction model and obtaining a concentration prediction sequence of the target pollutant in a future preset time window; the instruction generation module is used for obtaining a control instruction sequence of the target equipment by taking the concentration prediction sequence as input, taking a target concentration threshold value at a position and minimized energy consumption as constraint targets and solving through a preset optimization algorithm, and driving the target equipment based on the control instruction sequence.
8. A computer device comprising a memory and a processor, the memory storing a computer program, characterized in that the processor implements the steps of the method of any of claims 1 to 6 when the computer program is executed.
9. A computer readable storage medium, on which a computer program is stored, characterized in that the computer program, when being executed by a processor, implements the steps of the method of any of claims 1 to 6.
10. A computer program product comprising a computer program, characterized in that the computer program, when being executed by a processor, implements the steps of the method of any of claims 1 to 6.

Description

Big data-based air purifier driving control method, device, equipment, medium and program product Technical Field The application relates to the technical field of big data, in particular to an air purifier driving control method, device, equipment, medium and program product based on big data. Background With the acceleration of industrialization progress and the continuous improvement of urban level, the air pollution problem is increasingly prominent, and the air pollution problem becomes a global public health challenge. In this context, air purifiers have been widely used and continuously developed as a key device for improving indoor air quality. The traditional air purifier mainly removes particulate matters, gaseous pollutants and microorganisms in the air through physical filtration (such as a HEPA filter screen), electrostatic adsorption, photocatalysis, anion generation and other technical means. The core aim is to effectively reduce the concentration of PM2.5, pollen, peculiar smell, volatile Organic Compounds (VOCs), bacteria and viruses and other pollutants in a closed or semi-closed space, and ensure the health and comfort of residents. From the drive and control aspects, the operation mode of the existing air purifier has undergone an evolution from manual control to automatic intelligent control. Early products rely on the manual switching on and shutting down of user and adjust wind speed gear more, lack the real-time response ability to air quality. With the progress of sensor technology, modern air purifier generally integrates particulate matter sensor, gas sensor, temperature and humidity sensor etc. and can automatically detect ambient air quality and adjust running state accordingly, for example, automatically increase fan rotational speed when pollutant concentration rises, realizing "detection-feedback-regulation" closed-loop control. In addition, many products have supported remote control, enabling users to remotely monitor air quality data and manipulate the device by connecting to a smart phone application through Wi-Fi or bluetooth. The intelligent upgrading obviously improves the user experience and the purification efficiency, but is still limited to single machine automation of the equipment, and lacks higher-level coordination and optimization capability. In the related art, the rise of big data, the internet of things and artificial intelligence technology brings new opportunities for the development of air purifiers. The air purifier is connected to the Internet of things platform, interconnection and intercommunication among devices and data aggregation can be realized, and a foundation is laid for intelligent analysis based on big data. However, the driving control method of the current air purifier has the following technical problems: the current driving control method of the air purifier depends on a preset fixed threshold value or a simple rule, and has limited optimization in the aspects of overall energy efficiency and purification effect and needs to be optimized. Disclosure of Invention In view of the foregoing, it is desirable to provide an air cleaner driving control method, apparatus, computer device, computer readable storage medium, and computer program product based on big data that can enhance the energy efficiency and the cleaning efficiency of the air cleaner. In a first aspect, the present application provides a big data based air purifier driving control method. The method comprises the following steps: Acquiring real-time sensing data of target equipment and running state data of associated equipment, and acquiring external environment time sequence data through a preset data interface; preprocessing the real-time sensing data, the real-time running state data and the external environment time sequence data, extracting time sequence feature vectors and context feature vectors related to air quality dynamic change, wherein the preprocessing comprises data cleaning, data alignment and normalization processing; processing the time sequence feature vector and the context feature vector based on a pre-constructed air quality prediction model to obtain a concentration prediction sequence of the target pollutant in a future preset time window; and solving by a preset optimization algorithm by taking the concentration prediction sequence as input and taking a target concentration threshold value at a position and minimized energy consumption as constraint targets to obtain a control instruction sequence of the target equipment, and driving the target equipment based on the control instruction sequence. In one embodiment, the obtaining the control instruction sequence of the target device by taking the concentration prediction sequence as input, taking a target concentration threshold value at a bit and minimized energy consumption as constraint targets, and solving through a preset optimization algorithm includes: Constructing a multi-objective cost functio