Pressure signal identification and processing method

Abstract

The invention discloses a pressure signal identification and processing method, which relates to the technical field of sensor pressure detection, and is characterized in that pressure sensors are arranged for a plurality of industrial production equipment to acquire equipment operation pressure, and data processing nodes are constructed to process the equipment operation pressure of the industrial production equipment so as to generate electric pressure signals; constructing a signal identification network to input an electric pressure signal of the industrial production equipment, extracting a signal key feature corresponding to the electric pressure signal by the signal identification network, performing mode identification on the signal key feature to obtain a pressure situation of the industrial production equipment corresponding to the electric pressure signal, and constructing a pressure signal identification model; and acquiring an electric pressure signal of the industrial production equipment needing to be detected, inputting the electric pressure signal into the pressure signal identification model so as to acquire a pressure range of the industrial production equipment, performing early warning on the industrial production equipment of which the pressure range is not within a safe numerical value range at a display end, and arranging an inspector to overhaul the corresponding pressure sensor and the industrial production equipment by an administrator.

Description

Technical Field

[0001] The present invention relates to the technical field of sensor pressure detection, and in particular to a pressure signal recognition and processing method. Background Art

[0002] In the industrial production process, the operating status of equipment is directly related to production efficiency and safety. The operating pressure of industrial production equipment is one of the important parameters to measure the operating status of equipment. Traditional pressure monitoring methods often rely on manual inspections or simple sensor monitoring, which have problems such as poor real-time performance, weak data processing capabilities, and imperfect early warning mechanisms.

[0003] Therefore, how to achieve real-time and accurate identification and processing of pressure signals of industrial production equipment has become a problem that needs to be solved urgently. Summary of the invention

[0004] In order to solve the above problems, an object of the present invention is to provide a pressure signal recognition and processing method.

[0005] The purpose of the present invention can be achieved by the following technical solution: A pressure signal recognition and processing method comprises the following steps:

[0006] Step S1: arranging a plurality of pressure sensors at different industrial production equipment, and using the pressure sensors to collect the equipment operation pressure corresponding to the industrial production equipment, constructing a plurality of data processing nodes, each of which processes the equipment operation pressure of the industrial production equipment, and then generates a corresponding electric pressure signal;

[0007] Step S2: construct a signal recognition network, input the electric pressure signal corresponding to each industrial production equipment into the signal recognition network, and use the signal recognition network to extract the key signal features corresponding to each electric pressure signal;

[0008] Step S3: Perform pattern recognition on all key features of the signal, and then obtain the pressure situation of the industrial production equipment corresponding to the corresponding electric pressure signal;

[0009] Step S4: construct a pressure signal recognition model, obtain the electric pressure signal corresponding to the industrial production equipment to be tested, and input it into the pressure signal recognition model, thereby obtaining the pressure range of the corresponding industrial production equipment;

[0010] Step S5: The industrial production equipment whose pressure range is not within the safe value range will be warned on the display end, and the administrator will arrange inspection personnel to inspect and repair the corresponding pressure sensors and industrial production equipment.

[0011] Furthermore, a process of arranging a plurality of pressure sensors at different industrial production equipment and using the pressure sensors to collect the equipment operation pressure corresponding to the industrial production equipment includes:

[0012] A number of pressure sensors are numbered and recorded as i, i=1, 2, 3, ..., n, where n is a natural number greater than 0, and starting from i=1, working parameters are configured for each pressure sensor, the working parameters including pressure sensitivity, pressure value measurement range, response time and measurement accuracy;

[0013] Pressure sensors are arranged for several industrial production equipment to collect the equipment operation pressure of each industrial production equipment. There is a one-to-one mapping relationship between the pressure sensor and the industrial production equipment. The equipment operation pressure of the industrial production equipment collected by the pressure sensor numbered i is recorded as Pure[i].

[0014] Furthermore, a number of data processing nodes are constructed, each of which processes the equipment operation pressure of the industrial production equipment, and the process of generating a corresponding electric pressure signal includes:

[0015] Construct several data processing nodes and number them as j, j = 1, 2, 3, ..., m, where m is a natural number greater than 0, and construct a data mapping relationship between the data processing nodes and the industrial production equipment. The data mapping relationship is recorded as R, then R = <(i, Pure[i]), j>;

[0016] The processing performed by the data processing node includes signal conversion, signal amplification, and signal filtering;

[0017] Convert the equipment operating pressure of industrial production equipment into electrical pressure signal through signal conversion;

[0018] Select a signal amplification circuit to amplify the electric pressure signal, determine whether the signal amplification circuit has a fault, and determine whether the signal strength of the current electric pressure signal meets the standard, and perform different operations according to the determination results;

[0019] Select different types of filters for industrial production equipment to filter the voltage pressure signal. Different types of filters include high-pass filters, low-pass filters, and band-pass filters.

[0020] After completing signal conversion, signal amplification and signal filtering, the equipment operating pressure corresponding to each industrial production equipment generates its own final electric pressure signal.

[0021] Furthermore, a signal recognition network is constructed, and the electric pressure signal corresponding to each industrial production equipment is input into the signal recognition network. The process of extracting the key signal features corresponding to each electric pressure signal by the signal recognition network includes:

[0022] A signal recognition network is constructed by using convolutional neural network technology. The signal recognition network is composed of several signal recognition units, each of which is used to perform feature extraction operations on the electric pressure signal of an industrial production equipment;

[0023] According to the signal quantity of the electric pressure signal recorded in each signal recognition unit, the signal recognition network allocates the computing power resources corresponding to each signal recognition unit, and synchronously executes the feature extraction operations of several signal recognition units, thereby extracting the key signal features of each electric pressure signal. The key signal features include signal time domain features, signal frequency domain features and signal time-frequency features.

[0024] Furthermore, the process of performing pattern recognition on all key signal features and obtaining the pressure situation of industrial production equipment corresponding to the corresponding electric pressure signal includes:

[0025] Setting the ideal parameter sets of the signal time domain features, signal frequency domain features and signal time-frequency features in the signal key features, and performing pattern recognition by judging whether the signal time domain features, signal frequency domain features and signal time-frequency features meet their respective ideal parameter sets;

[0026] If any one of the signal time domain feature, the signal frequency domain feature, and the signal time-frequency feature does not conform to their respective ideal parameter sets, the electric pressure signal corresponding to the current signal key feature is judged to be an abnormal signal, and the pressure situation of the industrial production equipment corresponding to the abnormal signal is an abnormal situation;

[0027] If the signal time domain characteristics, signal frequency domain characteristics and signal time-frequency characteristics respectively meet their respective ideal parameter sets, the electric pressure signal corresponding to the current signal key characteristics is judged to be a normal signal, and the pressure situation of the industrial production equipment corresponding to the normal signal is a normal situation.

[0028] Furthermore, a pressure signal recognition model is constructed to obtain an electric pressure signal corresponding to the industrial production equipment to be tested, and input the electric pressure signal into the pressure signal recognition model, thereby obtaining the pressure range of the corresponding industrial production equipment. The process includes:

[0029] Construct a pressure signal recognition model, including a model input end and a model output end, input an electric pressure signal corresponding to an industrial production equipment with a known pressure situation at the model input end, and output the pressure range of the industrial production equipment corresponding to the corresponding electric pressure signal at the model output end;

[0030] If the pressure range output by the model output terminal meets the actual pressure range of the corresponding pressure situation of the industrial production equipment, then an effective recognition behavior of the pressure signal recognition model is recorded once, otherwise, an invalid recognition behavior is recorded once, and the model recognition rate corresponding to the pressure signal recognition model is generated according to the number of effective recognition behaviors and invalid recognition behaviors recorded by the pressure signal recognition model;

[0031] A model optimization threshold is set, and based on the numerical relationship between the model recognition rate and the model optimization threshold, it is determined whether the pressure signal recognition model needs to be trained, thereby generating the final pressure signal recognition model. The electric pressure signal of the industrial production equipment to be tested is input into the pressure signal recognition model, and the pressure signal recognition model outputs the pressure range of the equipment operating pressure of the industrial production equipment corresponding to the input electric pressure signal.

[0032] Furthermore, the process of determining whether the pressure signal recognition model needs to be trained includes:

[0033] The model recognition rate of the pressure signal recognition model is denoted as η;

[0034] ;

[0035] in, The number of valid recognition behaviors recorded for the pressure signal recognition model, The number of invalid recognition behaviors recorded for the pressure signal recognition model;

[0036] The model optimization threshold is denoted as M, M>0;

[0037] When η ≥ M, no operation is performed;

[0038] When η<M, a training set is set to perform model training on the pressure signal recognition model until η≥M is satisfied, at which time the model training is stopped.

[0039] Furthermore, the process of giving an early warning on the display terminal for industrial production equipment whose pressure range is not within the safe value range and the administrator arranging inspection personnel to inspect the corresponding pressure sensors and industrial production equipment includes:

[0040] Set a safe numerical range for the operating pressure of each industrial production equipment, compare the pressure range of each industrial production equipment obtained by the pressure signal recognition model with the respective safe numerical range, highlight the equipment number of the industrial production equipment whose pressure range is not within the safe numerical range on the display end, and generate corresponding early warning prompts;

[0041] The administrator retrieves all highlighted equipment numbers and locates all abnormal industrial production equipment. The administrator arranges inspection personnel to inspect and repair each abnormal industrial production equipment and the corresponding pressure sensors at the industrial production equipment.

[0042] Compared with the prior art, the present invention has the following beneficial effects:

[0043] 1. By arranging pressure sensors for each industrial production equipment, the equipment operation pressure of each industrial production equipment is collected in real time, and the equipment operation pressure is processed into a corresponding electric pressure signal by building a data processing node. A signal recognition network is built to extract the key signal features of the electric pressure signal. After pattern recognition of the key signal features, the corresponding pressure situation of the industrial production equipment is obtained, which effectively improves the accuracy of electric pressure signal recognition and the data processing capability of electric pressure signals, and timely diagnoses industrial production equipment with abnormal pressure situations.

[0044] 2. Construct a pressure signal recognition model, obtain the electric pressure signal corresponding to the industrial production equipment that needs to be tested and input it into the pressure signal recognition model to obtain the pressure range of the industrial production equipment. The industrial production equipment whose pressure range is not within the safe value range will be warned on the display end, and the administrator will arrange inspection personnel to inspect and repair the corresponding pressure sensors and industrial production equipment, providing a complete early warning method, effectively preventing the further expansion of equipment failures and safety accidents, and ensuring that industrial production equipment is in a safe operating state. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] Figure 1 It is a flow chart of the present invention. DETAILED DESCRIPTION

[0046] like Figure 1 As shown, a pressure signal recognition and processing method includes the following steps:

[0047] Step S1: arranging a plurality of pressure sensors at different industrial production equipment, and using the pressure sensors to collect the equipment operation pressure corresponding to the industrial production equipment, constructing a plurality of data processing nodes, each of which processes the equipment operation pressure of the industrial production equipment, and then generates a corresponding electric pressure signal;

[0048] Step S2: construct a signal recognition network, input the electric pressure signal corresponding to each industrial production equipment into the signal recognition network, and use the signal recognition network to extract the key signal features corresponding to each electric pressure signal;

[0049] Step S3: Perform pattern recognition on all key features of the signal, and then obtain the pressure situation of the industrial production equipment corresponding to the corresponding electric pressure signal;

[0050] Step S4: construct a pressure signal recognition model, obtain the electric pressure signal corresponding to the industrial production equipment to be tested, and input it into the pressure signal recognition model, thereby obtaining the pressure range of the corresponding industrial production equipment;

[0051] Step S5: The industrial production equipment whose pressure range is not within the safe value range will be warned on the display end, and the administrator will arrange inspection personnel to inspect and repair the corresponding pressure sensors and industrial production equipment.

[0052] It should be further explained that, in a specific implementation process, a plurality of pressure sensors are arranged at different industrial production equipment, and the process of using the pressure sensors to collect the equipment operation pressure corresponding to the industrial production equipment includes:

[0053] A number of pressure sensors are numbered and the number is recorded as i, i=1, 2, 3, ..., n, where n is a natural number greater than 0. Starting from i=1, corresponding working parameters are configured for each pressure sensor, and the working parameters include pressure sensitivity, pressure value measurement range, response time and measurement accuracy;

[0054] The pressure sensitivity refers to the ratio of the output pressure change of the pressure sensor to the input pressure change;

[0055] The pressure value measurement range refers to the maximum pressure value and the minimum pressure value that the pressure sensor can measure;

[0056] The response time refers to the time from when the pressure sensor receives the external input pressure to when it outputs a stable pressure value;

[0057] The measurement accuracy is used to indicate the degree of closeness between the pressure value measured by the pressure sensor and the actual pressure value;

[0058] Pressure sensors are arranged for several industrial production equipment. The equipment operating pressure corresponding to each industrial production equipment is collected through the pressure sensors. There is a one-to-one mapping relationship between the pressure sensors and the industrial production equipment. The equipment operating pressure of the industrial production equipment collected by the pressure sensor numbered i is recorded as Pure[i].

[0059] It should be further explained that, in the specific implementation process, a number of data processing nodes are constructed, each of which processes the equipment operation pressure of the industrial production equipment, and then generates a corresponding electric pressure signal, including:

[0060] Construct a number of data processing nodes, and number the data processing nodes, record the number as j, then j = 1, 2, 3, ..., m, where m is a natural number greater than 0, and construct a data mapping relationship between the data processing nodes and the industrial production equipment;

[0061] Let the data mapping relationship be R, then R = <(i, Pure[i]), j>;

[0062] The meaning of R = <(i, Pure[i]), j> is as follows:

[0063] The equipment operating pressure collected by the pressure sensor of the industrial production equipment numbered i is processed by the data processing node numbered j;

[0064] The processing of the equipment operation pressure of the industrial production equipment by the data processing node includes signal conversion, signal amplification and signal filtering;

[0065] The equipment operating pressure corresponding to the industrial production equipment is converted into a corresponding electric pressure signal through signal conversion;

[0066] Select a signal amplification circuit, amplify the electric pressure signal through the signal amplification circuit, obtain the signal strength before and after the signal amplification circuit amplifies the electric pressure signal, and record the signal strength before amplification as QD 前 , the signal intensity after amplification is recorded as QD 后 ;

[0067] Set the intensity increase threshold of the signal amplification circuit to amplify the electric pressure signal, and record it as QD 增幅 ;

[0068] If |QD 前 -QD 后 |<QD 增幅 , it is judged that the signal amplifier circuit is faulty and the signal amplifier circuit is repaired. If |QD 前 -QD 后 |≥QD 增幅 , then perform the following judgment operations:

[0069] Set the threshold of the electric pressure signal strength and record it as QD 达标 ;

[0070] When QD 后 ≥QD 达标 When the signal strength of the current electric pressure signal reaches the standard;

[0071] When QD<QD 达标 When the signal strength of the current electric pressure signal is judged to be not up to the standard;

[0072] If it is determined that the signal strength of the electric pressure signal does not meet the standard, the signal amplification circuit is replaced to re-amplify the current electric pressure signal. If it is determined that the signal strength of the electric pressure signal meets the standard, no operation is performed;

[0073] According to the production characteristics of industrial production equipment, select the corresponding type of filter to filter the electric pressure signal. Different types of filters include high-pass filters, low-pass filters and band-pass filters.

[0074] The low-pass filter allows the electric pressure signal with a frequency lower than the preset cut-off frequency F1 to pass through, and attenuates the electric pressure signal with a frequency greater than or equal to the cut-off frequency F1;

[0075] It should be noted that if we only focus on slowly changing electrical pressure signals, such as changes in electrical pressure signals within a few minutes or hours, we can use a low-pass filter to filter out rapidly changing high-frequency noise, such as noise caused by mechanical vibration in industrial production equipment, which uses mechanical power.

[0076] The high-pass filter allows the electric pressure signal with a frequency greater than or equal to the preset cut-off frequency F2 to pass through, and attenuates the electric pressure signal with a frequency less than the cut-off frequency F2;

[0077] It should be noted that if we only focus on the rapidly changing part of the electric-pressure signal, and the electric-pressure signal includes slowly changing signal components, such as the baseline drift of the electric-pressure signal caused by temperature drift, then a high-pass filter can be used to filter out the slowly changing signal components, for example, industrial production equipment is optoelectronic equipment.

[0078] The bandpass filter allows the electric pressure signal within the preset frequency range to pass through and attenuates the electric pressure signal outside the frequency range. The frequency range is set to (fa, fb), 0<fa<fb;

[0079] It should be noted that if you only focus on the changes in the electric pressure signal within a specific frequency range, for example, in hydraulic equipment, you may only focus on the pulsation of a specific frequency. At this time, the bandpass filter can be used to extract the electric pressure signal in this frequency range and filter out the electric pressure signal outside the frequency range.

[0080] After completing signal conversion, signal amplification and signal filtering, the equipment operating pressure corresponding to each industrial production equipment generates its own final electric pressure signal.

[0081] It should be further explained that, in the specific implementation process, a signal recognition network is constructed, and the electric pressure signal corresponding to each industrial production equipment is input into the signal recognition network. The signal recognition network extracts the signal key features corresponding to each electric pressure signal, and the process includes:

[0082] A signal recognition network is constructed by using convolutional neural network technology. The signal recognition network is composed of several signal recognition units, each of which is used to perform feature extraction operations on the electric pressure signal of an industrial production equipment;

[0083] The electric pressure signal of each industrial production equipment is recorded into the signal recognition network, and the signal recognition network assigns a signal recognition unit that has not yet performed a feature extraction operation to the recorded electric pressure signal until several signal recognition units have recorded the electric pressure signal;

[0084] According to the signal quantity of the electric pressure signal recorded in each signal recognition unit, the signal recognition network allocates the computing power resources corresponding to each signal recognition unit to ensure that the computing power resources can complete the feature extraction operation of the signal recognition unit on the electric pressure signal;

[0085] Synchronously perform feature extraction operations of a plurality of signal recognition units, thereby extracting key signal features corresponding to each electric pressure signal, wherein the key signal features include signal time domain features, signal frequency domain features, and signal time-frequency features;

[0086] The signal recognition unit uses the mean, variance, standard deviation, root mean square value, waveform factor, peak factor, margin factor, pulse factor, signal skewness and signal kurtosis of the statistical electric pressure signal as the signal time domain features;

[0087] The signal recognition unit converts the electric pressure signal into a frequency domain signal through Fourier transform, analyzes the frequency components of the frequency domain signal, and then extracts the signal frequency domain features, which include spectrum energy distribution, main frequency and spectrum entropy;

[0088] The signal recognition unit uses short-time Fourier transform or wavelet transform to extract the signal time-frequency characteristics of the electric pressure signal. The signal time-frequency characteristics are used to reflect the changes in the frequency components of the electric pressure signal at different time nodes, including but not limited to time-frequency ridges, time-frequency moments, rearranged spectra and marginal spectra.

[0089] It should be further explained that, in the specific implementation process, the process of pattern recognition of all key signal features and obtaining the pressure situation of industrial production equipment corresponding to the corresponding electric pressure signal includes:

[0090] Setting the ideal parameter sets of the signal time domain features, signal frequency domain features and signal time-frequency features in the signal key features, and performing pattern recognition by judging whether the signal time domain features, signal frequency domain features and signal time-frequency features meet their respective ideal parameter sets;

[0091] The content of pattern recognition is as follows:

[0092] If any one of the signal time domain feature, the signal frequency domain feature, and the signal time-frequency feature does not conform to their respective ideal parameter sets, the electric pressure signal corresponding to the current signal key feature is judged to be an abnormal signal, and the pressure situation of the industrial production equipment corresponding to the abnormal signal is an abnormal situation;

[0093] If the signal time domain characteristics, signal frequency domain characteristics and signal time-frequency characteristics meet their respective ideal parameter sets, then the electric pressure signal corresponding to the current signal key characteristics is judged to be a normal signal, and the pressure situation of the industrial production equipment corresponding to the normal signal is a normal situation;

[0094] It should be noted that the ideal parameter sets of the signal time domain features, signal frequency domain features and signal time-frequency features in the signal key features are generated by collecting the equipment operating pressure of the industrial production equipment that has been detected to be in normal operating state, processing the corresponding equipment operating pressure, and extracting the corresponding signal key features from the electric pressure signal. The signal key features extracted from the industrial production equipment in normal operating state are used as the feature data of the signal time domain features, signal frequency domain features and signal time-frequency features under ideal conditions.

[0095] It should be further explained that, in the specific implementation process, the process of constructing a pressure signal recognition model, obtaining the electric pressure signal corresponding to the industrial production equipment to be detected, and inputting it into the pressure signal recognition model, and then obtaining the pressure range of the corresponding industrial production equipment includes:

[0096] Construct a pressure signal recognition model, the pressure signal recognition model includes a model input end and a model output end, the model input end inputs an electric pressure signal corresponding to an industrial production equipment with a known pressure situation, and the model output end outputs a pressure range of the industrial production equipment corresponding to the corresponding electric pressure signal;

[0097] It should be noted that for industrial production equipment that has been tested, it corresponds to a pressure situation, that is, industrial production equipment with a known pressure situation. The pressure situation is represented by a corresponding pressure range. The pressure range at this time is the actual pressure range corresponding to the industrial production equipment. A pressure signal recognition model is now constructed to perform pressure range detection on industrial production equipment with a known pressure situation, and the pressure range detected corresponding to the industrial production equipment with a known pressure situation is compared with the actual pressure range under real conditions to determine the model recognition rate of the current pressure signal recognition model; that is, the corresponding data of the industrial production equipment with a known pressure situation is used as a test set for the pressure signal recognition model to establish a pressure signal recognition model that meets the requirements.

[0098] If the pressure range output by the model output end to the industrial production equipment meets the actual pressure range of the corresponding pressure situation of the industrial production equipment, then an effective recognition behavior of the pressure signal recognition model is recorded;

[0099] If the pressure range output by the model output end to the industrial production equipment does not conform to the actual pressure range of the industrial production equipment corresponding to the pressure situation, an invalid recognition behavior of the pressure signal recognition model is recorded;

[0100] According to the number of effective recognition behaviors and invalid recognition behaviors recorded by the pressure signal recognition model, the model recognition rate corresponding to the pressure signal recognition model is generated, and the model recognition rate is recorded as η. The formula of the model recognition rate is as follows:

[0101] ;

[0102] in, The number of valid recognition behaviors recorded for the pressure signal recognition model, The number of invalid recognition behaviors recorded for the pressure signal recognition model;

[0103] Set the model optimization threshold and record it as M, M>0;

[0104] When η ≥ M, no operation is performed;

[0105] When η<M, a training set is set to train the pressure signal recognition model until η≥M is satisfied, and then the model training is stopped;

[0106] The electric pressure signal of the industrial production equipment to be tested is input into the pressure signal recognition model, and the pressure signal recognition model outputs the pressure range of the equipment operating pressure of the industrial production equipment corresponding to the input electric pressure signal.

[0107] It should be further explained that, in the specific implementation process, the industrial production equipment whose pressure range is not within the safe value range will be warned on the display end, and the administrator will arrange inspection personnel to inspect the corresponding pressure sensors and industrial production equipment. The process includes:

[0108] Set a safe numerical range for the operating pressure of each industrial production equipment, compare the pressure range of each industrial production equipment obtained by the pressure signal recognition model with the respective safe numerical range, highlight the equipment number of the industrial production equipment whose pressure range is not within the safe numerical range on the display end, and generate corresponding early warning prompts;

[0109] The administrator retrieves all highlighted equipment numbers and locates all abnormal industrial production equipment. The administrator arranges inspection personnel to inspect and repair each abnormal industrial production equipment and the corresponding pressure sensors at the industrial production equipment.

[0110] The above embodiments are only used to illustrate the technical method of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical method of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical method of the present invention.

Claims

1. A method for identifying and processing a pressure signal, characterized in that: The following steps are involved: Step S1: arranging a plurality of pressure sensors at different industrial production equipment, and using the pressure sensors to collect the equipment operation pressure corresponding to the industrial production equipment, constructing a plurality of data processing nodes, each of which processes the equipment operation pressure of the industrial production equipment, and then generates a corresponding electric pressure signal; Step S2: construct a signal recognition network, input the electric pressure signal corresponding to each industrial production equipment into the signal recognition network, and use the signal recognition network to extract the key signal features corresponding to each electric pressure signal; Step S3: Perform pattern recognition on all key features of the signal, and then obtain the pressure situation of the industrial production equipment corresponding to the corresponding electric pressure signal; Step S4: construct a pressure signal recognition model, obtain the electric pressure signal corresponding to the industrial production equipment to be tested, and input it into the pressure signal recognition model, thereby obtaining the pressure range of the corresponding industrial production equipment; Step S5: The industrial production equipment whose pressure range is not within the safe value range will be warned on the display end, and the administrator will arrange inspection personnel to inspect and repair the corresponding pressure sensors and industrial production equipment.

2. A pressure signal recognition and processing method according to claim 1, characterized in that: The process of arranging a number of pressure sensors at different industrial production equipment and using the pressure sensors to collect the equipment operating pressure corresponding to the industrial production equipment includes: A number of pressure sensors are numbered and recorded as i, i=1, 2, 3, ..., n, where n is a natural number greater than 0, and starting from i=1, working parameters are configured for each pressure sensor, the working parameters including pressure sensitivity, pressure value measurement range, response time and measurement accuracy; Pressure sensors are arranged for several industrial production equipment to collect the equipment operation pressure of each industrial production equipment. There is a one-to-one mapping relationship between the pressure sensor and the industrial production equipment. The equipment operation pressure of the industrial production equipment collected by the pressure sensor numbered i is recorded as Pure[i].

3. A pressure signal recognition and processing method according to claim 2, characterized in that: The process of constructing several data processing nodes, each of which processes the equipment operation pressure of industrial production equipment and generates corresponding electrical pressure signals includes: Construct several data processing nodes and number them as j, j = 1, 2, 3, ..., m, where m is a natural number greater than 0, and construct a data mapping relationship between the data processing nodes and the industrial production equipment. The data mapping relationship is recorded as R, then R = <(i, Pure[i]), j>; The processing performed by the data processing node includes signal conversion, signal amplification, and signal filtering; Convert the equipment operating pressure of industrial production equipment into electrical pressure signal through signal conversion; Select a signal amplification circuit to amplify the electric pressure signal, determine whether the signal amplification circuit has a fault, and determine whether the signal strength of the current electric pressure signal meets the standard, and perform different operations according to the determination results; Select different types of filters for industrial production equipment to filter the voltage pressure signal. Different types of filters include high-pass filters, low-pass filters, and band-pass filters. After completing signal conversion, signal amplification and signal filtering, the equipment operating pressure corresponding to each industrial production equipment generates its own final electric pressure signal.

4. A pressure signal recognition and processing method according to claim 3, characterized in that: Construct a signal recognition network, input the electric pressure signal corresponding to each industrial production equipment into the signal recognition network, and extract the key signal features corresponding to each electric pressure signal by the signal recognition network. The process includes: A signal recognition network is constructed by using convolutional neural network technology. The signal recognition network is composed of several signal recognition units, each of which is used to perform feature extraction operations on the electric pressure signal of an industrial production equipment; According to the signal quantity of the electric pressure signal recorded in each signal recognition unit, the signal recognition network allocates the computing power resources corresponding to each signal recognition unit, and synchronously executes the feature extraction operations of several signal recognition units, thereby extracting the key signal features of each electric pressure signal. The key signal features include signal time domain features, signal frequency domain features and signal time-frequency features.

5. A pressure signal recognition and processing method according to claim 4, characterized in that: The process of pattern recognition of all key signal features and obtaining the corresponding pressure situation of industrial production equipment corresponding to the corresponding electric pressure signal includes: Setting the ideal parameter sets of the signal time domain features, signal frequency domain features and signal time-frequency features in the signal key features, and performing pattern recognition by judging whether the signal time domain features, signal frequency domain features and signal time-frequency features meet their respective ideal parameter sets; If any one of the signal time domain feature, the signal frequency domain feature, and the signal time-frequency feature does not conform to their respective ideal parameter sets, the electric pressure signal corresponding to the current signal key feature is judged to be an abnormal signal, and the pressure situation of the industrial production equipment corresponding to the abnormal signal is an abnormal situation; If the signal time domain characteristics, signal frequency domain characteristics and signal time-frequency characteristics respectively meet their respective ideal parameter sets, the electric pressure signal corresponding to the current signal key characteristics is judged to be a normal signal, and the pressure situation of the industrial production equipment corresponding to the normal signal is a normal situation.

6. A pressure signal recognition and processing method according to claim 5, characterized in that: The process of constructing a pressure signal recognition model, obtaining the electric pressure signal corresponding to the industrial production equipment to be tested, and inputting it into the pressure signal recognition model, and then obtaining the pressure range of the corresponding industrial production equipment includes: Construct a pressure signal recognition model, including a model input end and a model output end, input an electric pressure signal corresponding to an industrial production equipment with a known pressure situation at the model input end, and output the pressure range of the industrial production equipment corresponding to the corresponding electric pressure signal at the model output end; If the pressure range output by the model output terminal meets the actual pressure range of the corresponding pressure situation of the industrial production equipment, then an effective recognition behavior of the pressure signal recognition model is recorded once, otherwise, an invalid recognition behavior is recorded once, and the model recognition rate corresponding to the pressure signal recognition model is generated according to the number of effective recognition behaviors and invalid recognition behaviors recorded by the pressure signal recognition model; A model optimization threshold is set, and based on the numerical relationship between the model recognition rate and the model optimization threshold, it is determined whether the pressure signal recognition model needs to be trained, thereby generating the final pressure signal recognition model. The electric pressure signal of the industrial production equipment to be tested is input into the pressure signal recognition model, and the pressure signal recognition model outputs the pressure range of the equipment operating pressure of the industrial production equipment corresponding to the input electric pressure signal.

7. A pressure signal recognition and processing method according to claim 6, characterized in that: The process of determining whether the pressure signal recognition model needs to be trained includes: The model recognition rate of the pressure signal recognition model is denoted as η; ； in, The number of valid recognition behaviors recorded for the pressure signal recognition model, The number of invalid recognition behaviors recorded for the pressure signal recognition model; The model optimization threshold is denoted as M, M>0; When η ≥ M, no operation is performed; When η<M, a training set is set to perform model training on the pressure signal recognition model until η≥M is satisfied, at which time the model training is stopped.

8. A pressure signal recognition and processing method according to claim 7, characterized in that: The process of giving early warning on the display end to industrial production equipment whose pressure range is not within the safe value range and the administrator arranging inspection personnel to inspect the corresponding pressure sensors and industrial production equipment includes: Set a safe numerical range for the operating pressure of each industrial production equipment, compare the pressure range of each industrial production equipment obtained by the pressure signal recognition model with the respective safe numerical range, highlight the equipment number of the industrial production equipment whose pressure range is not within the safe numerical range on the display end, and generate corresponding early warning prompts; The administrator retrieves all highlighted equipment numbers and locates all abnormal industrial production equipment. The administrator arranges inspection personnel to inspect and repair each abnormal industrial production equipment and the corresponding pressure sensors at the industrial production equipment.

Images