Method and device for adjusting rotating speed, nonvolatile storage medium and electronic equipment

Abstract

The invention discloses a rotating speed adjusting method and device, a nonvolatile storage medium and electronic equipment. The method comprises the steps that environmental parameters of a to-be-detected area at multiple detection moments are obtained, and the environmental parameters are parameters related to the dust concentration of the to-be-detected area; a dust concentration prediction model is adopted to extract and process high-order characteristics of the environmental parameters, a dust concentration prediction value is obtained, the dust concentration prediction value is used for indicating the dust concentration of the to-be-detected area at the next detection moment, and the high-order characteristics are time sequence characteristics of the environmental parameters related to time; the rotating speed of the target equipment is adjusted according to the dust concentration predicted value, and when the rotating speed of the target equipment is different, the exhaust intensity of the to-be-detected area is different. The technical problem that the equipment loss is increased due to inaccurate adjustment results caused by manual adjustment of the rotating speed of the equipment with the air circulation function in related technologies is solved.

Description

Technical Field

[0001] The present application relates to the field of automation control technology, and in particular to a method and device for adjusting rotation speed, a non-volatile storage medium, and an electronic device. Background Art

[0002] Duct blowers are one of the important equipment used in steel production. At production sites such as steelmaking and coke ovens, the concentration of metal, coal dust and other dust in the air is relatively high. Working in this environment for a long time will cause serious harm to the health of on-site personnel and pollute the environment around the factory park. The use of duct blowers can discharge the generated dust, exhaust gas, etc. out of the factory along the air duct, reduce the dust concentration in the on-site environment, protect the health and safety of on-site personnel, and at the same time process and recycle useful materials for reproduction. In related technologies, the rotation speed of equipment with air circulation functions such as duct blowers is mainly controlled by on-site personnel according to the dust conditions in the on-site environment. There are problems such as the inability to adjust the rotation speed of such equipment in a timely and accurate manner, waste of resources, and increased equipment wear.

[0003] To address the above-mentioned problems, no effective solution has been proposed yet. Summary of the invention

[0004] The embodiments of the present application provide a method and device for adjusting the rotation speed, a non-volatile storage medium, and an electronic device, so as to at least solve the technical problem of increased equipment loss due to inaccurate adjustment results caused by manual adjustment of the rotation speed of equipment with air circulation function in the related art.

[0005] According to one aspect of an embodiment of the present application, a method for adjusting the rotation speed is provided, including: obtaining environmental parameters of an area to be detected at multiple detection moments, wherein the environmental parameters are parameters related to the dust concentration of the area to be detected, and the difference between any two adjacent detection moments in the multiple detection moments is a preset value; using a dust concentration prediction model to extract and process high-order features of the environmental parameters to obtain a dust concentration prediction value, wherein the dust concentration prediction value is used to indicate the dust concentration of the area to be detected at the next detection moment, and the high-order features are time series features related to the environmental parameters and time, and the dust concentration prediction model is obtained by training a neural network model using historical environmental parameters of the area to be detected as training data; adjusting the rotation speed of a target device according to the dust concentration prediction value, wherein when the rotation speed of the target device is different, the exhaust intensity of the area to be detected is different, and the target device is a device with an air circulation function in the area to be detected.

[0006] Optionally, a dust concentration prediction model is used to extract and process high-order features of environmental parameters to obtain a dust concentration prediction value, including: determining the detection time corresponding to each environmental parameter; arranging multiple different types of environmental parameters corresponding to the same detection time to obtain multiple environmental parameter sequences corresponding to multiple detection times, wherein the types of environmental parameters include: humidity, temperature, air pressure, air flow rate, and dust concentration; performing convolution processing on multiple environmental parameter sequences at different neural network layers of the dust concentration prediction model to obtain high-order features, wherein the first neural network layer is used to extract high-order features, and other neural network layers are used to process high-order features, and different neural network layers use convolution kernels of different dimensions; and determining the dust concentration prediction value based on the processing results of the high-order features and multiple environmental parameter sequences.

[0007] Optionally, convolution processing is performed on multiple environmental parameter sequences in different neural network layers of the dust concentration prediction model to obtain high-order features, including: for each neural network layer, determining the detection time corresponding to each environmental parameter sequence, and arranging the multiple environmental parameter sequences into a matrix to be processed in order of the detection time from early to late, wherein each row of the matrix to be processed is an environmental parameter sequence; for the Nth row of the matrix to be processed, using a preset convolution kernel to perform convolution processing on a new matrix composed of data of the first N rows of the matrix to be processed to obtain a first convolution result, wherein the preset convolution kernel is a matrix with the same dimension as the new matrix; determining the sum of multiple first convolution results as the convolution sum, and processing the preset convolution kernel and multiple convolution sums to obtain high-order features.

[0008] Optionally, the preset convolution kernel and multiple first convolution results are processed to obtain high-order features, including: filling the preset convolution kernel to obtain multiple other convolution kernels of different dimensions; using another convolution kernel to perform convolution processing on the convolution sum in the target other neural network layer to obtain a second convolution result, wherein the target other neural network layer is the other neural network layer connected to the first neural network layer; according to the connection order of the remaining other neural network layers, using other convolution kernels of different dimensions in different other neural network layers to perform convolution processing on the input data to obtain the target convolution result output by the last other neural network layer, wherein, during the convolution processing, the input of the first other neural network in sequence is the second convolution result, and the output of the previous other neural network layer is used as the input data of the next other neural network layer; the high-order features are determined according to the target convolution result and the environmental parameter sequence.

[0009] Optionally, high-order features are processed, including: obtaining hidden features of the dust concentration detection model, wherein the hidden features are used to indicate the relationship between the environmental parameters and the detection time, and the hidden features are determined by the dust concentration prediction model according to the training data during the training process; using a recurrent neural memory network included in the dust concentration detection model to perform feature fusion processing on the hidden features and the high-order features to obtain feature fusion results, wherein the recurrent neural memory network includes a plurality of sequentially connected recurrent units, and during the feature fusion process, the output of the previous recurrent unit is used as the input of the next recurrent unit; converting the feature fusion result into a target data sequence, wherein the amount of data included in the target data sequence is the same as the amount of data included in the environmental parameter sequence, and the dust concentration value included in the target data sequence is the dust concentration prediction value.

[0010] Optionally, adjusting the rotational speed of the target device according to the dust concentration prediction value includes: obtaining a historical prediction value of the dust concentration corresponding to the previous detection moment, and determining a prediction value of a control parameter according to the historical prediction value of the dust concentration and the dust concentration prediction value, wherein the control parameter is a parameter used to control the rotational speed of the target device; obtaining multiple historical deviation values ​​corresponding to multiple historical detection moments before the current detection moment, and determining a prediction deviation value according to the multiple historical deviation values, wherein the historical deviation value is the difference between the historical prediction value of the dust concentration and the historical measurement value of the dust concentration; determining a target value of the control parameter according to the prediction value and the prediction deviation value, and adjusting the value of the control parameter to the target value.

[0011] Optionally, before using the dust concentration prediction model to extract and process high-order features of environmental parameters, the method includes: classifying environmental parameters of the same type into a data set; for each data set, arranging the environmental parameters into a sequence to be processed in order of detection time from early to late; for each sequence to be processed, determining a value range according to the values ​​of the environmental parameters contained in the sequence to be processed; when the value of the environmental parameter does not fall within the value range, performing numerical correction processing on the environmental parameter.

[0012] According to another aspect of an embodiment of the present application, a device for adjusting the rotational speed is also provided, including: an acquisition module, used to acquire environmental parameters of an area to be detected at multiple detection moments, wherein the environmental parameters are parameters related to the dust concentration of the area to be detected, and the difference between any two adjacent detection moments in the multiple detection moments is a preset value; a processing module, used to extract and process high-order features of the environmental parameters using a dust concentration prediction model to obtain a dust concentration prediction value, wherein the dust concentration prediction value is used to indicate the dust concentration of the area to be detected at the next detection moment, and the high-order features are time series features related to the environmental parameters and time, and the dust concentration prediction model is obtained by training a neural network model using the historical environmental parameters of the area to be detected as training data; an adjustment module, used to adjust the rotational speed of a target device according to the dust concentration prediction value, wherein the target device is a device with air circulation function in the area to be detected.

[0013] According to another aspect of an embodiment of the present application, a non-volatile storage medium is provided, in which a computer program is stored, wherein the method for adjusting the rotation speed is executed by running the computer program in a device where the non-volatile storage medium is located.

[0014] According to another aspect of an embodiment of the present application, an electronic device is provided, including a memory and a processor, wherein a computer program is stored in the memory, and the processor is configured to execute the above-mentioned method for adjusting the rotation speed through the computer program.

[0015] According to another aspect of an embodiment of the present application, a computer program product is also provided, including computer instructions, and when the computer instructions are executed by a processor, the above-mentioned method for adjusting the rotation speed is implemented.

[0016] In an embodiment of the present application, environmental parameters of the area to be detected at multiple detection moments are obtained, wherein the environmental parameters are parameters related to the dust concentration of the area to be detected, and the difference between any two adjacent detection moments in the multiple detection moments is a preset value; a dust concentration prediction model is used to extract and process high-order features of the environmental parameters to obtain a dust concentration prediction value, wherein the dust concentration prediction value is used to indicate the dust concentration of the area to be detected at the next detection moment, and the high-order feature is a time series feature related to the environmental parameter and time. The dust concentration prediction model is obtained by training a neural network model using the historical environmental parameters of the area to be detected as training data; the rotation speed of the target device is adjusted according to the dust concentration prediction value, wherein when the rotation speed of the target device is different, the rotation speed of the area to be detected is adjusted according to the dust concentration prediction value. The exhaust intensity is different, and the target device is a device with air circulation function in the area to be detected. The environmental parameters of the area where the equipment with air circulation function (such as duct machine) is located are collected and input into the trained neural network model to predict the dust concentration value at the current moment, and further determine the rotation speed of the equipment with air circulation function (such as duct machine) at the current moment according to the dust concentration value. The purpose of applying the neural network model to predict dust concentration and adjusting the rotation speed of the duct machine is achieved, thereby achieving the technical effect of improving the accuracy of speed control, reducing equipment wear and saving energy, and further solving the technical problem of inaccurate adjustment result and increased equipment loss caused by manual adjustment of the rotation speed of the equipment with air circulation function in the related technology. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] The drawings described herein are used to provide a further understanding of the present application and constitute a part of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0018] Figure 1 is a hardware structure block diagram of a computer terminal for implementing a method for adjusting a rotation speed according to an embodiment of the present application;

[0019] Figure 2 is a flowchart of a method for adjusting the rotation speed according to an embodiment of the present application;

[0020] Figure 3 1 is a schematic diagram of a data processing flow of a temporal convolutional network (TCN) according to an embodiment of the present application;

[0021] Figure 4 is a schematic diagram of a data processing flow of an improved cycle unit according to an embodiment of the present application;

[0022] Figure 5 is a structural diagram of a device for adjusting the rotation speed according to an embodiment of the present application;

[0023] Figure 6 is a schematic diagram of a system for adjusting rotation speed according to an embodiment of the present application;

[0024] Figure 7 It is the indicator comparison result when multiple models are used to process the same test set according to the embodiment of the present application;

[0025] Figure 8 It is the prediction result of dust concentration predicted by multiple models according to the embodiments of the present application based on the same validation set. DETAILED DESCRIPTION

[0026] In order to enable those skilled in the art to better understand the solution of the present application, the technical solution 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 only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present application.

[0027] It should be noted that the terms "first", "second", etc. in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments of the present application described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device comprising a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0028] In order to better understand the embodiments of the present application, the technical terms involved in the embodiments of the present application are explained as follows:

[0029] Proportional-integral-differential (PID) control algorithm: an algorithm that controls according to the proportion (P), integral (I) and differential (D) of the deviation.

[0030] In the related art, when the neural network model is applied to the scenario of predicting dust concentration, there are problems such as inaccurate extraction of features leading to inaccurate dust concentration prediction results, and high energy consumption of running the neural network model due to the complex structure of the neural network. In order to solve this problem, the embodiments of the present application provide relevant solutions, which are described in detail below.

[0031] According to an embodiment of the present application, an embodiment of a method for adjusting the rotational speed is provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and although a logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0032] The method embodiments provided in the embodiments of the present application can be executed in a mobile terminal, a computer terminal or a similar computing device. Figure 1 FIG. 1 shows a hardware structure block diagram of a computer terminal for implementing a method for adjusting the rotation speed. Figure 1 As shown, the computer terminal 10 may include one or more (102a, 102b, ..., 102n are used to illustrate) processors 102 (the processor 102 may include but is not limited to a processing device such as a microprocessor MCU or a programmable logic device FPGA), a memory 104 for storing data, and a transmission device 106 for communication functions. In addition, it may also include: a display, an input / output interface (I / O interface), a universal serial bus (USB) port (which may be included as one of the ports of the BUS bus), a network interface, a power supply and / or a camera. It can be understood by those skilled in the art that Figure 1 The structure shown is only for illustration and does not limit the structure of the above electronic device. Figure 1 More or fewer components as shown, or with Figure 1 Different configurations are shown.

[0033] It should be noted that the one or more processors 102 and / or other data processing circuits described above may generally be referred to herein as "data processing circuits". The data processing circuits may be embodied in whole or in part as software, hardware, firmware, or any other combination thereof. In addition, the data processing circuit may be a single independent processing module, or may be incorporated in whole or in part into any of the other components in the computer terminal 10. As described in the embodiments of the present application, the data processing circuit acts as a processor control (e.g., selection of a variable resistor terminal path connected to an interface).

[0034] The memory 104 can be used to store software programs and modules of application software, such as the program instructions / data storage device corresponding to the method for adjusting the rotation speed in the embodiment of the present application. The processor 102 executes various functional applications and data processing by running the software programs and modules stored in the memory 104, that is, the above-mentioned method for adjusting the rotation speed is realized. The memory 104 may include a high-speed random access memory, and may also include a non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some examples, the memory 104 may further include a memory remotely arranged relative to the processor 102, and these remote memories may be connected to the computer terminal 10 via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0035] The transmission device 106 is used to receive or send data via a network. The specific example of the above network may include a wireless network provided by a communication provider of the computer terminal 10. In one example, the transmission device 106 includes a network adapter (Network Interface Controller, NIC), which can be connected to other network devices through a base station so as to communicate with the Internet. In one example, the transmission device 106 can be a radio frequency (RF) module, which is used to communicate with the Internet wirelessly.

[0036] The display may be, for example, a touch screen liquid crystal display (LCD) that enables a user to interact with a user interface of the computer terminal 10 .

[0037] The embodiment of the present application provides a method for adjusting the rotation speed which can be operated in the above operating environment. Figure 2 is a flowchart of the steps of the method for adjusting the rotation speed provided in an embodiment of the present application, such as Figure 2 As shown, the method comprises the following steps:

[0038] Step S202, obtaining environmental parameters of the area to be detected at multiple detection times, wherein the environmental parameters are parameters related to the dust concentration in the area to be detected, and the difference between any two adjacent detection times in the multiple detection times is a preset value.

[0039] An embodiment of the present application provides a control system for predicting rotational speed based on multi-source information fusion, wherein the multi-source information is environmental information of an area (i.e., an area to be detected) where a device with air circulation function controlled by rotational speed is located. Therefore, in step S202, environmental parameters related to dust concentration in the area to be detected are first obtained, that is, the obtained environmental parameters can affect the dust concentration in the area to be detected. Since the dust concentration in the area changes in real time with the environment of the area, in this embodiment, the predicted rotational speed is updated by periodically obtaining environmental parameters in the area to be detected; for example, according to a preset detection cycle (for example, 6 minutes as a cycle), environmental parameters of the area to be detected are collected at multiple detection times belonging to the detection cycle; wherein the difference between any two adjacent detection times is equal, and is the preset duration of the detection cycle (i.e., the preset value).

[0040] Step S204, using the dust concentration prediction model to extract and process the high-order features of the environmental parameters to obtain a dust concentration prediction value, wherein the dust concentration prediction value is used to indicate the dust concentration of the area to be detected at the next detection moment, the high-order features are the time series features related to the environmental parameters and time, and the dust concentration prediction model is obtained by training the neural network model using the historical environmental parameters of the area to be detected as training data.

[0041] Since the speed is predicted based on environmental parameters, after the environmental parameters of the area to be detected are obtained in step S202, in step S204, the collected environmental parameters are used as the input of the dust concentration prediction model, and the dust concentration prediction model performs feature extraction processing on the input environmental parameters to obtain high-order features for representing the relationship between dust concentration and time, and calculates the dust concentration prediction value based on the high-order features, wherein the dust concentration prediction value is the predicted dust concentration of the area to be detected at the next detection moment. As can be seen from the above, since the high-order features represent the relationship between dust concentration and time, they have time series characteristics, which are actually time-related time series features extracted from environmental parameters, for example, a vector for representing the change area of ​​environmental parameters over time; and the reason why the dust concentration prediction model can be used to predict the dust concentration of the area to be detected at the next detection moment is that the dust concentration prediction model is a neural network model trained with the environmental parameters of the area to be detected collected at multiple historical detection moments before the current detection moment (or the previous detection moment) as training data, and the relationship between dust concentration, environmental parameters and time is learned during the training process.

[0042] In this embodiment, when the dust concentration prediction model is used, the dust concentration prediction model can be loaded into the memory. For example, the raw data of the dust concentration prediction model can be loaded from the non-volatile memory into the volatile memory so that the processor runs the dust concentration prediction model. The raw data of the dust concentration prediction model refers to unprocessed data, which usually includes parameters and structural data of the dust concentration prediction model. The structural data can be a calculation relationship based on the parameters, such as the forward propagation calculation relationship between the intermediate layers and the neurons. Specifically, the structural data can include code related to the structure of the dust concentration prediction model, such as code for performing related calculations between the intermediate layers and the neurons.

[0043] In one embodiment, a memory may be divided for loading the dust concentration prediction model, which may include a structure data storage area and a parameter storage area. The structure data storage area is used to store structure-related codes, and the parameters referenced by it can point to the address of a specific parameter in the parameter storage area through a pointer. During the training process of the dust concentration prediction model, it may be necessary to frequently update the parameters, and the parameter values ​​in the parameter storage area can be updated.

[0044] According to an optional embodiment of the present application, a dust concentration prediction model is used to extract and process high-order features of environmental parameters to obtain a dust concentration prediction value, including: determining the detection time corresponding to each environmental parameter; arranging multiple different types of environmental parameters corresponding to the same detection time to obtain multiple environmental parameter sequences corresponding to multiple detection times, wherein the types of environmental parameters include: humidity, temperature, air pressure, air flow rate, and dust concentration; performing convolution processing on multiple environmental parameter sequences at different neural network layers of the dust concentration prediction model to obtain high-order features, wherein the first neural network layer is used to extract high-order features, and other neural network layers are used to process high-order features, and different neural network layers use convolution kernels of different dimensions; and determining the dust concentration prediction value according to the processing results of the high-order features and multiple environmental parameter sequences.

[0045] The scene that needs to be tested for dust concentration (i.e., the area to be tested) usually includes a variety of scenes such as steel production sites. In such scenes, the parameters related to dust concentration include: temperature, humidity, air pressure, air flow rate and dust concentration, etc.; wherein, such parameters can be obtained by temperature sensors, humidity sensors, air pressure sensors, wind speed sensors, dust concentration detectors, etc. that are pre-set in the area to be tested and communicate with the device / system running the dust concentration prediction model; such parameters change with the production process, that is, the environmental parameters and dust concentrations at different times are different, that is, the dust concentration is related to time. To predict the dust concentration value at the next detection time based on the environmental parameters, it is necessary to pre-process the input data to make it related to time. In this embodiment, the dust concentration prediction model first divides the input environmental parameters into multiple sets according to the detection time when they are collected, and composes multiple environmental parameters obtained at the same detection time into an environmental parameter sequence. Therefore, multiple environmental parameter sequences can be obtained, wherein the time when each environmental parameter is collected is the corresponding detection time. The dust concentration prediction model is a neural network model, which includes multiple neural network layers. The dust concentration prediction model performs convolution processing on the environmental parameter sequence generated by the above arrangement in different neural network layers, and can extract the time-related time series features (i.e., high-order features) in the environmental parameters, wherein different convolution kernels are used when performing convolution processing in different neural network layers; in this embodiment, the dust concentration prediction model uses the first neural network layer to extract high-order features, and the remaining other neural network layers process the extracted high-order features. The dust concentration prediction model further predicts the dust concentration value of the area to be detected at the next detection moment based on the processing results of the high-order features and the input environmental parameter sequence, and outputs the dust concentration prediction value.

[0046] Optionally, convolution processing is performed on multiple environmental parameter sequences in different neural network layers of the dust concentration prediction model to obtain high-order features, including: for each neural network layer, determining the detection time corresponding to each environmental parameter sequence, and arranging the multiple environmental parameter sequences into a matrix to be processed in order of the detection time from early to late, wherein each row of the matrix to be processed is an environmental parameter sequence; for the Nth row of the matrix to be processed, using a preset convolution kernel to perform convolution processing on a new matrix composed of data of the first N rows of the matrix to be processed to obtain a first convolution result, wherein the preset convolution kernel is a matrix with the same dimension as the new matrix; determining the sum of multiple first convolution results as the convolution sum, and processing the preset convolution kernel and multiple convolution sums to obtain high-order features.

[0047] Figure 3 This is a schematic diagram of the data processing flow of TCN, such as Figure 3As shown, in this embodiment, when the dust concentration prediction model extracts high-order features, first, multiple environmental parameter sequences are arranged in order from early to late according to their corresponding detection times to obtain a matrix to be processed, wherein each environmental parameter sequence is used as row data of the matrix to be processed, and each column of the matrix to be processed contains multiple environmental parameters obtained at different detection times; for example, the environmental parameter sequence includes: X0, X1, X2, X3, X4, wherein the detection time corresponding to X0 is the earliest, the detection time corresponding to X1 is later than the detection time corresponding to X0, the detection time corresponding to X2 is later than the detection time corresponding to X1, the detection time corresponding to X3 is later than the detection time corresponding to X2, and the detection time corresponding to X4 is the latest, then the matrix to be processed (X) can be expressed as follows: X = {X0, X1, X2, X3, X4}, each environmental parameter sequence X k(k＝0、1、2、3、4) ={x M}, where M represents M environmental parameters; each column X of the matrix to be processed l(k＝0,…M) ={x N}, where N represents N detection moments; x N is the same environmental parameter collected at the Nth moment. The dust concentration prediction model uses causal convolution and dilated convolution to extract high-order features. The first neural network layer is the temporal convolutional network (TCN) layer, which contains multiple layers of TCN. The multiple layers of TCN extract high-order features of the input data in parallel to achieve rapid modeling of dust concentration changes. Specifically, when TCN is used for causal convolution, only the environmental parameters at multiple moments before the prediction moment are considered without involving future parameters. Since the data at one moment (time point) is a row of the matrix to be processed, when extracting high-order features in each layer of the temporal convolutional network, in order to extract the temporal characteristics between the environmental parameters and time at each time point, it is necessary to work together based on the environmental parameter sequences at multiple time points before this time point. Specifically, when processing the matrix to be processed, when extracting each row of When the timing features are obtained, the first N rows of this row need to work together; in this embodiment, the first N rows of the matrix to be processed are regarded as a new matrix. When extracting the timing features of each row, a preset convolution kernel of the same dimension as the new matrix is ​​used to convolve the new matrix to obtain a (first) convolution result; for example, when predicting the timing features of the time point corresponding to the third row of the matrix to be processed, the environmental parameter sequence of the first three rows of the matrix to be processed is used; through the above method, a matrix to be processed containing N rows can be convolved to obtain N (first) convolution results, and the sum of these N (first) convolution results (i.e., the convolution sum) and the preset convolution kernel can be processed together to obtain high-order features. The above causal convolution process can be expressed as a formula: Among them, X is the matrix to be processed, W is the parameter matrix containing multiple preset convolution kernels, each preset convolution kernel is also a matrix, x i represents the environmental parameters at the Ith detection time; k represents the kth element, for example, w k represents the kth element in the parameter matrix, x k Represents the kth element in the matrix to be processed, and N represents the dimension of the convolution kernel, which is also the dimension of the matrix to be processed.

[0048] Optionally, the preset convolution kernel and multiple first convolution results are processed to obtain high-order features, including: filling the preset convolution kernel to obtain multiple other convolution kernels of different dimensions; using another convolution kernel to perform convolution processing on the convolution sum in the target other neural network layer to obtain a second convolution result, wherein the target other neural network layer is the other neural network layer connected to the first neural network layer; according to the connection order of the remaining other neural network layers, using other convolution kernels of different dimensions in different other neural network layers to perform convolution processing on the input data to obtain the target convolution result output by the last other neural network layer, wherein, during the convolution processing, the input of the first other neural network in sequence is the second convolution result, and the output of the previous other neural network layer is used as the input data of the next other neural network layer; the high-order feature target convolution result is determined as the high-order feature according to the target convolution result and the environmental parameter sequence.

[0049] Still Figure 3As shown, in this embodiment, multiple (first) convolution results and preset convolution kernels outputted from the first layer of TCN are processed together in other neural network layers of the dust concentration prediction model to obtain high-order features. Specifically, the dimension of the preset convolution kernel is changed by filling the preset convolution kernel to obtain multiple new convolution kernels (i.e., other convolution kernels) of different dimensions. In this embodiment, in order to avoid affecting the accuracy of the prediction results, the preset convolution kernel can be filled with the number "0". After filling to obtain multiple new convolution kernels (i.e., other convolution kernels) of different dimensions, the new convolution kernels (i.e., other convolution kernels) of different dimensions and the input data of the layer are convolved in other (target) neural network layers connected to the neural network layer (i.e., the first neural network layer) for extracting high-order features in the dust concentration prediction model, respectively. The input data of each target neural network layer is the output of the previous target neural network layer connected to it. In this embodiment, the first layer of TCN outputs the sum of multiple (first) convolution results (i.e., the convolution sum), then the neural network layer connected to the first layer of TCN (i.e., the target other neural network layer) uses a new convolution kernel with the same dimension as the convolution sum to perform convolution processing on the convolution sum, and uses the obtained (second) convolution result as the input of the next neural network layer connected to the target other neural network layer; for example, the preset convolution kernels with expansion coefficients of 2 and 4 can be used in the second and third layers of TCN for dilated convolution respectively. In this embodiment, the above dilated convolution method is adopted to increase the coverage of the input data time series by increasing the convolution kernel, which can better extract the long-term dependencies in the time series and solve the problem of limited receptive field of the original network structure. The dilated convolution process in this embodiment can be expressed as a formula: Among them, X is the convolution sum, is in the form of a matrix, W is a matrix composed of multiple new convolution kernels, w k is the kth element in W, i.e., the kth new convolution kernel, representing the number of new convolution kernels, s represents the step size of the new convolution kernel, i.e., the interval at which the new convolution kernel slides on the input data, and d represents the expansion factor (expansion coefficient); in the above manner, the convolution kernel at time k is convolved with the environmental parameter (sequence) at time [k-(Ns)d], which can better extract the long-term dependency in the time series. After the dust concentration prediction model uses the above method to perform multi-layer processing on the convolution and, the last layer of TCN will output a final convolution result (i.e., the target convolution result). In this embodiment, the target convolution result and the input environmental parameter sequence are residually connected to obtain high-order features; for example, if TCN has 3 layers in total, the result output by the third layer of TCN (target convolution result) is subjected to residual operation with the input data (environmental parameter sequence) at the same time. Take time t 31 For example, the environmental parameter sequence x at this moment 31After the 1×1 convolution branch, and the third layer output x after the linear rectification function (ReLU) activation and regularization (Dropout) layer 31 ″, perform element-by-element addition operations to obtain the final output result of the TCN network (i.e., high-order features); this step alleviates the problem of gradient vanishing or gradient exploding due to the increase in depth of the model.

[0050] According to another optional embodiment of the present application, high-order features are processed, including: obtaining hidden features of the dust concentration detection model, wherein the hidden features are used to indicate the relationship between the environmental parameters and the detection time, and the hidden features are determined by the dust concentration prediction model according to the training data during the training process; using a recurrent neural memory network included in the dust concentration detection model to perform feature fusion processing on the hidden features and the high-order features to obtain feature fusion results, wherein the recurrent neural memory network includes a plurality of sequentially connected recurrent units, and during the feature fusion process, the output of the previous recurrent unit is used as the input of the next recurrent unit; converting the feature fusion result into a target data sequence, wherein the amount of data included in the target data sequence is the same as the amount of data included in the environmental parameter sequence, and the dust concentration value included in the target data sequence is the dust concentration prediction value.

[0051] Figure 4 It is a schematic diagram of the data processing flow of the improved cyclic unit in the dust concentration prediction model. The method provided in the embodiment of the present application also adds a feature fusion operation in the long short-term memory network (Long Short Term Memory, LSTM) unit in the dust concentration prediction model. Specifically, before the operation of the cyclic unit gating structure of the LSTM unit, the multiplication operation of the unit hidden state and the input state is added. Specifically, in this embodiment, when the high-order features are processed in the LSTM unit, the hidden features and the high-order features are subjected to feature fusion operation in the recurrent neural memory network; wherein, the hidden features are the features extracted by the dust concentration prediction model according to the training data during the training process, which are used to indicate the correlation between the environmental parameters and the detection time (for example, the hidden features can be a feature vector used to indicate the changing trend of the environmental parameters with the change of the detection time). Figure 4 As shown, the feature fusion process can be expressed by the following formula: Among them, X t represents the environmental parameter sequence input at time t, H t-1 represents the hidden state (hidden feature) at time (t-1), σ represents the activation function; feature fusion result Used to update input status, feature fusion results Used to update the hidden state. After a feature fusion is performed through the above method, the updated hidden state can be obtained. and the updated input state In addition, according to the schematic diagram of the above formula, it can be seen that the structure of the recurrent unit is used in the feature fusion process, and the output of the previous recurrent unit is used as the input of the next recurrent unit connected to it. The semantic correlation between the input data at the current moment and the hidden state of the previous unit output is enhanced through the above feature fusion, so as to avoid the recurrent unit from forgetting important features and improve the accuracy of the prediction result. In this embodiment, when the hidden state and the input state are updated during feature fusion, the activation function σ is used to update the updated hidden state Updated input status A linear memory cell unit in LSTM (C t-1 ), multiple different convolution kernels (w f 、w i 、w g 、w o ) are fused, and the fusion process is as follows:

[0053] Among them, f t is the forget gate of the recurrent unit, i t is the input gate of the recurrent unit, g t is the input modulation gate of the recurrent unit, o t is the output gate of the recurrent unit, b f 、b i 、b c 、b o is a pre-set bias of different values. Tanh is an activation function. Through the above steps, when the features are fused, f t Forgetting unimportant features in memory cells makes i t and g t The important features of the current input state of the learning unit are updated according to the result of adding the above two parts, and then normalized with o after tanh activation function. t Operation to obtain output results Among them, the generated The process is shown in the following formula: t =f t ⊙C t-1 +i t ⊙g t , Among them, C t is the updated memory cell unit. Since the LSTM in the embodiment of the present application adopts the self-attention mechanism, the gate structure operation outputs After that, three different convolution kernels (w q 、wk 、w v ) to obtain three vectors of the self-attention mechanism: Query(q t )、Key(k t ) and Value(V t ); When the self-attention mechanism is used for calculation, q t With k t The similarity score is obtained by multiplication, and the weight coefficient matrix is ​​obtained after scaling and normalization by the normalized exponential function (Softmax), and then added to V t Do the multiplication to get the final output H of the recurrent unit t (That is, the result of feature fusion). In this step, the dust concentration prediction model enhances the ability to extract global features in long sequences by assigning weights, thereby improving the accuracy of dust concentration prediction. This step can be expressed as the following formula: When performing feature fusion, repeat the above steps, increase the network depth by stacking 2 layers of recurrent units, and improve the model's ability to extract high-order features of the input sequence. After obtaining the feature fusion result, in this embodiment, the feature fusion result is converted into a sequence by dimensionality reduction to obtain a target data sequence, wherein the amount of data contained in the sequence obtained by the feature fusion result conversion (i.e., the target data sequence) is the same as the amount of data contained in the environmental parameter sequence, that is, the target data sequence has the same dimension as the environmental parameter sequence. In fact, the target data sequence contains prediction values ​​of different types of environmental parameters, and the dust concentration prediction value is obtained by querying the set in the target data sequence. In the above embodiment, converting the feature fusion result into a target data sequence can be implemented in the fully connected layer (Dense layer) of the dust concentration prediction model. In the fully connected layer, the dimension of the feature fusion result is adjusted to 1 to obtain the final prediction sequence. The variable corresponding to the dust concentration in the prediction sequence is the value predicted by the dust concentration prediction model for the on-site environmental dust concentration at the next detection moment.

[0054] Step S206, adjusting the rotation speed of the target device according to the dust concentration prediction value, wherein when the rotation speed of the target device is different, the exhaust intensity of the area to be detected is different, and the target device is a device with air circulation function in the area to be detected.

[0055] After predicting the dust concentration value of the area to be detected at the next detection moment in step S204, in step S206, the rotation speed determined according to the dust concentration prediction value is the rotation speed of the device with air circulation function (i.e., the target device) in the area to be detected at the next detection moment, wherein the target device reduces the dust concentration in the area to be detected by exhausting air to the area to be detected, and when the target device operates at different rotation speeds, the intensity of exhaust air to the area to be detected is different; therefore, the technical effect of adjusting the dust concentration in the area to be detected can be achieved by adjusting the rotation speed of the target device.

[0056] Optionally, adjusting the rotational speed of the target device according to the dust concentration prediction value includes: obtaining a historical prediction value of the dust concentration corresponding to the previous detection moment, and determining a prediction value of a control parameter according to the historical prediction value of the dust concentration and the dust concentration prediction value, wherein the control parameter is a parameter used to control the rotational speed of the target device; obtaining multiple historical deviation values ​​corresponding to multiple historical detection moments before the current detection moment, and determining a prediction deviation value according to the multiple historical deviation values, wherein the historical deviation value is the difference between the historical prediction value of the dust concentration and the historical measurement value of the dust concentration; determining a target value of the control parameter according to the prediction value and the prediction deviation value, and adjusting the value of the control parameter to the target value.

[0057] After predicting the dust concentration of the area to be detected at the next detection time through the above embodiments, in this embodiment, the rotation speed of the target device is determined according to the predicted dust concentration. Specifically, the dust concentration prediction value at the latest moment is read, and a result is calculated using the PID control algorithm, and the rotation speed of the target device is adjusted using the result. In this embodiment, when adjusting the rotation speed of the target device according to the dust concentration prediction value, first, it is necessary to determine the predicted value u(t) of the control parameter of the rotation speed corresponding to the dust concentration prediction value c(t) according to the following formula: where the target device is currently rotating at the rotation speed corresponding to u(t),

[0058]

[0059] Where c(t-1) is the dust concentration prediction value of the area to be detected at the last detection time (i.e., the historical dust concentration prediction value), T is the sampling time (i.e., the detection time), and K p is the proportionality coefficient, T i is the integration coefficient, T dis the differential coefficient, i represents the i-th detection moment, and t is the total number of detection moments. Furthermore, in order to improve the accuracy of adjusting the rotational speed, the deviation value of the currently output parameter value (i.e., the predicted deviation value Δu(t)) is determined according to the deviation value between the dust concentration predicted value at the predicted moment (i.e., the current moment) and the preset dust concentration value, as well as the deviation value of the dust concentration at multiple historical detection moments close to the predicted moment (the current detection moment) (i.e., the historical deviation value), and the predicted deviation value Δu(t) is used to correct the rotational speed of the target device. In this embodiment, the deviation value of the previously output parameter value (i.e., the predicted deviation value Δu(t)) is determined according to the following formula,

[0060]

[0061] Among them, e(t) is the deviation value of the dust concentration prediction value at time t (i.e., prediction time / current time), e(t-1) is the deviation value of the dust concentration prediction value at the previous detection time (t-1) before the current time, and e(t-2) is the deviation value of the dust concentration prediction value at the two detection times before the current time (t-2), among which e(t-1) and e(t-2) are both historical deviation values. Finally, the sum of u(t) and Δu(t) is the value to which the control parameter of the target device is finally adjusted (i.e., the target value). The control parameter can be a current percentage, and Δu(t) can be a positive number or a negative number.

[0062] According to some optional embodiments of the present application, before using the dust concentration prediction model to extract and process the high-order features of the environmental parameters, it includes: classifying the environmental parameters of the same type into a data set; for each data set, arranging the environmental parameters into a sequence to be processed in order from early to late according to the detection time; for each sequence to be processed, determining a value range according to the values ​​of the environmental parameters contained in the sequence to be processed; when the value of the environmental parameter does not fall within the value range, performing numerical correction processing on the environmental parameter.

[0063] In the above embodiments, the collected environmental parameters are directly input into the dust concentration prediction model to predict the dust concentration at the next detection moment, which can also achieve the technical effect of adjusting the rotation speed of the target equipment. However, if the environmental parameters are preprocessed and then input into the dust concentration prediction model for prediction, the accuracy of the prediction results can be improved. Therefore, in some embodiments, the environmental parameters can be preprocessed as follows: a preset number of environmental parameters are selected from the acquired environmental parameters, and the selected environmental parameters are processed in a sliding window manner; for example, the latest 32 values ​​of five environmental parameters, namely temperature, humidity, air pressure, air flow rate and dust mass concentration, collected by sensors at the steel production site can be read from the database and subjected to sliding window processing. Before performing sliding window processing, the environmental parameters belonging to the same category among the selected environmental parameters are divided into a data set. For each data set, the detection time corresponding to each environmental parameter is arranged from early to late into a sequence (i.e., a sequence to be processed), and slides on each sequence to be processed according to a preset order. Each time the sliding occurs, multiple environmental parameters included in the preset window are selected, and a value range is determined based on the multiple environmental parameters. The environmental parameters whose values ​​among the multiple environmental parameters included in the preset window do not fall within the above-determined value range are corrected, wherein the preset window is the preset number of environmental parameters selected each time, and the value range can be based on the average value μ of the multiple environmental parameters. k and standard deviation σ k Determined jointly, for example, the value range can be (μ k -3σ k , μ k +3σ k ), the correction process can be to use the mean of multiple environmental parameters contained in the preset window to replace the environmental parameter whose value does not belong to the value range. Taking a certain category of environmental parameter x as an example, for time t n Collect value x n , select the previous n-1 consecutive historical observation values ​​to form the sequence X k ={x k1 ,x k2 ,…,x kn}, calculate its mean μ k and standard deviation σ k : If x n ∈(μ k -3σ k , μ k +3σ k ), then determine x n is a normal value. nIf it falls outside this interval, it is determined to be an outlier and denoised and interpolated. The mean of the observed values ​​at the previous five moments is selected as the current corrected observed value. All environmental parameters are preprocessed according to the above process, and then the data is normalized to the interval [0,1], and the preprocessing is completed.

[0064] Through the above steps, it is possible to utilize the dust concentration prediction model to extract the time dimension and high-order features of the input environmental parameters by means of causal convolution and dilated convolution operations, thereby solving the problems of low feature extraction efficiency and limited receptive field of the existing network model; by adding the multiplication operation of the unit hidden state and the input state in the prediction process of the dust concentration prediction model, it is possible to avoid the problem of forgetting or failing to extract important features of dust concentration changes due to the weak contextual semantic correlation between the two in time sequence in the gated operation of the cyclic unit, thereby improving the accuracy of the prediction results; by adding the step of assigning weights to the operation results after the operation of the cyclic gating structure, the dust concentration prediction model is improved in its ability to extract global features of dust concentration changes, thereby further improving the accuracy of the prediction results.

[0065] Figure 5 is a structural diagram of a device for adjusting the rotation speed provided in an embodiment of the present application, such as Figure 5 As shown, the device includes: an acquisition module 50, which is used to obtain environmental parameters of the area to be detected at multiple detection moments, wherein the environmental parameters are parameters related to the dust concentration of the area to be detected, and the difference between any two adjacent detection moments in the multiple detection moments is a preset value; a processing module 52, which is used to extract and process high-order features of the environmental parameters using a dust concentration prediction model to obtain a dust concentration prediction value, wherein the dust concentration prediction value is used to indicate the dust concentration of the area to be detected at the next detection moment, and the high-order feature is a time series feature related to the environmental parameter and time. The dust concentration prediction model is obtained by training a neural network model using the historical environmental parameters of the area to be detected as training data; an adjustment module 54, which is used to adjust the rotation speed of the target device according to the dust concentration prediction value, wherein the target device is a device with air circulation function in the area to be detected.

[0066] It should be noted that Figure 5 The preferred implementation of the illustrated embodiment can be found in Figure 2 The relevant description of the illustrated embodiment will not be repeated here.

[0067] Figure 6 is a schematic diagram of a system for adjusting the rotation speed. The method provided in the embodiment of the present application can also be used for Figure 6 The system shown, Figure 6As shown in the figure, the system includes the following hardware architectures that can communicate: a data acquisition sensor group for collecting environmental parameters, a database server and a high-performance computing server for data processing, and a field automation control cabinet for controlling equipment with air circulation function; wherein, the data acquisition sensor group communicates with the database server through a 5G gateway, and the data acquisition sensor group includes: a temperature sensor, a humidity sensor, an air pressure sensor, a wind speed sensor, and a dust concentration detector; each sensor is deployed in the area to be detected where dust removal is required, and collects temperature, humidity, air pressure, wind speed and dust concentration data; each sensor is connected to a 5G gateway, and the 5G gateway has corresponding ports, reads sensor information, and sends it to a message broker server (MQTT Broker) at intervals of 6 minutes. The database server is deployed with a data subscription program and a database service, which subscribes to the collected data sent by the sensor to the MQTT Broker, and stores it in the database on the server after parsing. The high-performance computing server is deployed with a dust concentration prediction model and a speed mapping program. The high-performance computing server communicates with the automation control cabinet through the 5G gateway, and adjusts the speed of the duct machine (i.e., equipment with air circulation function) through the programmable logic controller (PLC). In this embodiment, the PID algorithm is used to determine the value of the speed-related control variable and output the current percentage according to the output increment. When the PID control algorithm is used, multiple parameters such as differential coefficient, integral coefficient and proportional coefficient are required. In this embodiment, the sampling time T in the PID algorithm is set to 180, the preset dust concentration value is set to 25, and the proportional coefficient (K p ) is set to 2.0, the integral coefficient (T i ) is set to 0.8, the differential coefficient (T d ) is set to 0.2. In this embodiment, the actual production data of the steel plant for three consecutive years, and various on-site environmental data collected at a frequency of 6 minutes are used to train the dust concentration prediction model, and the data training set, validation set and test set are divided into 60%, 20%, and 20% ratios. In the training stage, the input data is calculated to obtain the predicted value, and the loss function difference is calculated by comparing with the actual value, so as to continuously update the network parameters. The validation set is used to check whether the model is overfitting. In the testing stage, the prediction performance of the trained model is tested based on the test set. In this embodiment, the mean square error (MAE), root mean square error (RMSE) and determination coefficient (R-squared) are used to evaluate the accuracy of the model in predicting dust concentration. The calculation formulas for the three indicators are as follows Among them, m is the total number of samples, i is the i-th sample, and x i is the actual dust concentration value of the i-th sample, Predict the dust concentration value for the i-th sample model, is the average dust concentration of the samples. The lower the MAE and RMSE scores of the model, and the higher the R-Squared score, the better the prediction performance of the model.

[0068] Figure 7 It is the comparison result of indicators when multiple models are used to process the same test set. Figure 8 It is the prediction result of multiple models predicting dust concentration based on the same validation set. In order to verify the ability of the dust concentration prediction model provided in the embodiment of the present application in predicting dust concentration, the dust concentration prediction model in the embodiment of the present application, the long short-term memory network model combined with convolutional neural network (CNN-LSTM), the long and short-term time series network model (LSTNet), the recurrent neural network model combined with residual connection and jump connection (RSC-GRU) and the prediction model based on deep learning (DeepAR) are processed on the same test set for comparative experiments. The scores of various indicators of each model on the test set are as follows: Figure 7 As shown, according to Figure 7 It can be seen that the dust concentration prediction model provided in the embodiment of the present application can more accurately predict the dust concentration, and the speed adjustment of the duct fan based on the dust concentration prediction model provided in the embodiment of the present application is also more accurate. Figure 8 As shown, the dust concentration prediction values ​​at multiple detection moments output by multiple models are compared with the actual dust concentration values ​​(i.e., Ground Truth) at each detection moment included in the validation set. The dust concentration value predicted by the dust concentration prediction model (i.e., Proposed Model) provided in the embodiment of the present application is closer to the actual dust concentration value.

[0069] An embodiment of the present application further provides a non-volatile storage medium, in which a computer program is stored, wherein the above method for adjusting the rotation speed is executed by running the computer program in a device where the non-volatile storage medium is located.

[0070] The above-mentioned non-volatile storage medium is used to store programs that perform the following functions: obtaining environmental parameters of the area to be detected at multiple detection times, wherein the environmental parameters are parameters related to the dust concentration of the area to be detected, and the difference between any two adjacent detection times in the multiple detection times is a preset value; using a dust concentration prediction model to extract and process high-order features of the environmental parameters to obtain a dust concentration prediction value, wherein the dust concentration prediction value is used to indicate the dust concentration of the area to be detected at the next detection time, and the high-order features are time series features related to the environmental parameters and time. The dust concentration prediction model is obtained by training a neural network model using the historical environmental parameters of the area to be detected as training data; adjusting the rotation speed of the target device according to the dust concentration prediction value, wherein when the rotation speed of the target device is different, the exhaust intensity of the area to be detected is different, and the target device is a device with air circulation function in the area to be detected.

[0071] An embodiment of the present application further provides an electronic device, including a memory and a processor, wherein a computer program is stored in the memory, and the processor is configured to execute the above method for adjusting the rotation speed through the computer program.

[0072] The processor in the above-mentioned electronic device is used to run a program that performs the following functions: obtaining environmental parameters of the area to be detected at multiple detection times, wherein the environmental parameters are parameters related to the dust concentration of the area to be detected, and the difference between any two adjacent detection times in the multiple detection times is a preset value; using a dust concentration prediction model to extract and process high-order features of the environmental parameters to obtain a dust concentration prediction value, wherein the dust concentration prediction value is used to indicate the dust concentration of the area to be detected at the next detection time, and the high-order features are time series features related to the environmental parameters and time. The dust concentration prediction model is obtained by training a neural network model using the historical environmental parameters of the area to be detected as training data; adjusting the rotation speed of the target device according to the dust concentration prediction value, wherein when the rotation speed of the target device is different, the exhaust intensity of the area to be detected is different, and the target device is a device with air circulation function in the area to be detected.

[0073] An embodiment of the present application also provides a computer program product, including computer instructions, which implement the above method for adjusting the rotation speed when executed by a processor.

[0074] It should be noted that the various modules in the above-mentioned device for adjusting the speed can be program modules (for example, a set of program instructions for implementing a certain specific function) or hardware modules. For the latter, it can be expressed in the following forms, but is not limited to this: the expression form of each of the above-mentioned modules is a processor, or the functions of each of the above-mentioned modules are implemented by a processor.

[0075] The serial numbers of the above-mentioned embodiments of the present application are for description only and do not represent the advantages or disadvantages of the embodiments.

[0076] In the above embodiments of the present application, the description of each embodiment has its own emphasis. For parts that are not described in detail in a certain embodiment, please refer to the relevant description of other embodiments.

[0077] In the several embodiments provided in this application, it should be understood that the disclosed technical content can be implemented in other ways. Among them, the device embodiments described above are only schematic. For example, the division of the units can be a logical function division. There may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of units or modules, which can be electrical or other forms.

[0078] 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 units. Some or all of the units may be selected according to actual needs to achieve the purpose of the present embodiment.

[0079] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0080] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application, or the part that contributes to the relevant technology or all or part of the technical solution, can be embodied in the form of a software product, which is stored in a storage medium and includes several instructions for a computer device (which can be a personal computer, a server or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: U disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), mobile hard disk, disk or optical disk, etc. Various media that can store program codes.

[0081] The above is only a preferred implementation of the present application. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present application. These improvements and modifications should also be regarded as the scope of protection of the present application.

Claims

1. A method for adjusting the rotation speed, characterized in that: include: Acquire environmental parameters of the area to be detected at multiple detection times, wherein the environmental parameters are parameters related to the dust concentration of the area to be detected, and the difference between any two adjacent detection times in the multiple detection times is a preset value; A dust concentration prediction model is used to extract and process high-order features of the environmental parameters to obtain a dust concentration prediction value, wherein the dust concentration prediction value is used to indicate the dust concentration of the area to be detected at the next detection moment, the high-order features are time series features related to the environmental parameters, and the dust concentration prediction model is obtained by training a neural network model using historical environmental parameters of the area to be detected as training data; The rotation speed of the target device is adjusted according to the dust concentration prediction value, wherein when the rotation speed of the target device is different, the exhaust intensity for the area to be detected is different, and the target device is a device with air circulation function in the area to be detected.

2. The method according to claim 1, characterized in that The dust concentration prediction model is used to extract and process the high-order features of the environmental parameters to obtain the dust concentration prediction value, including: Determine the detection time corresponding to each of the environmental parameters; Arranging a plurality of different types of environmental parameters corresponding to the same detection moment to obtain a plurality of environmental parameter sequences corresponding to the plurality of detection moments, wherein the types of environmental parameters include: humidity, temperature, air pressure, air flow rate, and dust concentration; Convolution processing is performed on the multiple environmental parameter sequences in different neural network layers of the dust concentration prediction model to obtain the high-order features, wherein the first neural network layer is used to extract the high-order features, and the other neural network layers are used to process the high-order features, and different neural network layers use convolution kernels of different dimensions; The dust concentration prediction value is determined according to the processing result of the high-order feature and the plurality of environmental parameter sequences.

3. The method according to claim 2, characterized in that Convolution processing is performed on the multiple environmental parameter sequences at different neural network layers of the dust concentration prediction model to obtain the high-order features, including: For each of the neural network layers, determine the detection time corresponding to each of the environmental parameter sequences, and arrange the multiple environmental parameter sequences into a matrix to be processed in the order of the detection time from early to late, wherein each row of the matrix to be processed is one of the environmental parameter sequences; For the Nth row of the matrix to be processed, a preset convolution kernel is used to perform convolution processing on a new matrix composed of the data of the first N rows of the matrix to be processed to obtain a first convolution result, wherein the preset convolution kernel is a matrix with the same dimension as the new matrix; The sum of the multiple first convolution results is determined as a convolution sum, and the preset convolution kernel and the multiple convolution sums are processed to obtain the high-order feature.

4. The method according to claim 3, characterized in that Processing the preset convolution kernel and the plurality of the first convolution results to obtain the high-order features includes: Filling the preset convolution kernel to obtain multiple other convolution kernels with different dimensions; Using one of the other convolution kernels in the target other neural network layer to perform convolution processing on the convolution sum to obtain a second convolution result, wherein the target other neural network layer is the other neural network layer connected to the first neural network layer; According to the connection order of the remaining other neural network layers, the other convolution kernels of different dimensions are used in different other neural network layers to perform convolution processing on the input data, and the target convolution result output by the last other neural network layer is obtained, wherein, during the convolution processing, the input of the first other neural network is the second convolution result, and the output of the previous other neural network layer is used as the input data of the next other neural network layer; The high-order features are determined according to the target convolution result and the environmental parameter sequence.

5. The method according to claim 2, characterized in that: Processing the high-order features includes: Acquire a hidden feature of the dust concentration detection model, wherein the hidden feature is used to indicate the relationship between the environmental parameter and the detection time, and the hidden feature is determined by the dust concentration prediction model according to the training data during the training process; The hidden features and the high-order features are subjected to feature fusion processing by using a recurrent neural memory network included in the dust concentration detection model to obtain a feature fusion result, wherein the recurrent neural memory network includes a plurality of recurrent units connected in sequence, and during the feature fusion process, the output of the previous recurrent unit is used as the input of the next recurrent unit; The feature fusion result is converted into a target data sequence, wherein the amount of data contained in the target data sequence is the same as the amount of data contained in the environmental parameter sequence, and the dust concentration value contained in the target data sequence is the dust concentration prediction value.

6. The method according to claim 1, characterized in that Adjusting the rotation speed of the target device according to the dust concentration prediction value includes: Obtaining a historical predicted value of dust concentration corresponding to the last detection moment, and determining a predicted value of a control parameter according to the historical predicted value of dust concentration and the predicted value of dust concentration, wherein the control parameter is a parameter for controlling the rotation speed of the target device; Acquire multiple historical deviation values ​​corresponding to multiple historical detection moments before the current detection moment, and determine a predicted deviation value according to the multiple historical deviation values, wherein the historical deviation value is the difference between the historical predicted value of dust concentration and the historical measured value of dust concentration; A target value of the control parameter is determined according to the predicted value and the predicted deviation value, and the value of the control parameter is adjusted to the target value.

7. The method according to claim 1, characterized in that Before using the dust concentration prediction model to extract and process the high-order features of the environmental parameters, it includes: Classifying the environmental parameters of the same type into a data set; For each of the data sets, the environmental parameters are arranged in order from early to late according to the detection time into a sequence to be processed; For each of the sequences to be processed, determining a value range according to the values ​​of the environmental parameters included in the sequence to be processed; When the value of the environmental parameter does not fall within the value range, a value correction process is performed on the environmental parameter.

8. A device for adjusting the rotation speed, characterized in that: include: An acquisition module, used to acquire environmental parameters of the area to be detected at multiple detection moments, wherein the environmental parameters are parameters related to the dust concentration of the area to be detected, and the difference between any two adjacent detection moments in the multiple detection moments is a preset value; A processing module, used to extract and process the high-order features of the environmental parameters using a dust concentration prediction model to obtain a dust concentration prediction value, wherein the dust concentration prediction value is used to indicate the dust concentration of the area to be detected at the next detection moment, the high-order features are time series features related to the environmental parameters, and the dust concentration prediction model is obtained by training a neural network model using the historical environmental parameters of the area to be detected as training data; An adjustment module is used to adjust the rotation speed of the target device according to the dust concentration prediction value, wherein the target device is a device with air circulation function in the area to be detected.

9. A non-volatile storage medium, characterized in that: The non-volatile storage medium stores a computer program, wherein the method for adjusting the rotation speed according to any one of claims 1 to 7 is executed by running the computer program in the device where the non-volatile storage medium is located.

10. An electronic device comprising a memory and a processor, characterized in that: A computer program is stored in the memory, and the processor is configured to execute the method for adjusting the rotation speed according to any one of claims 1 to 7 through the computer program.

11. A computer program product comprising computer instructions, characterized in that: When the computer instructions are executed by a processor, the method for adjusting the rotation speed described in any one of claims 1 to 7 is implemented.

Images