CN119847875A

CN119847875A - Signal processing system, method and device based on computer

Info

Publication number: CN119847875A
Application number: CN202510314939.3A
Authority: CN
Inventors: 张灏
Original assignee: Beijing Asiainfo Data Co ltd
Current assignee: Beijing Asiainfo Data Co ltd
Priority date: 2025-03-18
Filing date: 2025-03-18
Publication date: 2025-04-18

Abstract

The invention relates to the technical field of computer signal processing, in particular to a signal processing system, a method and a device based on a computer, which comprises a high-precision acquisition module, a high-precision acquisition module and a high-precision acquisition module, wherein the high-precision acquisition module is used for acquiring signals; the intelligent monitoring system comprises a distributed processing module, a multi-signal fusion module, a real-time signal early warning module, an analog signal simulation module, an intelligent diagnosis and maintenance module, a computer simulation technology generation module and a system operation state monitoring and fault diagnosis module, wherein the distributed processing module is used for realizing distributed acquisition and processing of signals, the multi-signal fusion module is used for carrying out fusion processing on acquired signals of different types, the real-time signal early warning module is used for realizing real-time monitoring and early warning on abnormal signals, the analog signal simulation module is used for generating analog signals through the computer simulation technology to simulate a real environment, and the intelligent diagnosis and maintenance module is used for carrying out monitoring and fault diagnosis on the system operation state through a long-period memory network and providing maintenance reminding. The invention can realize high-precision acquisition of signals, meet the monitoring requirement in complex environments, improve the accuracy and reliability of signal processing and reduce the risk of device faults and maintenance cost.

Description

Signal processing system, method and device based on computer

Technical Field

The present invention relates to the field of computer signal processing technologies, and in particular, to a signal processing system, method and device based on a computer.

Background

In the field of modern industry, military and environmental monitoring, signal processing systems play a vital role. Conventional signal processing systems typically rely on a single sensor or centralized processing architecture, which have limitations in processing multi-source signals in complex environments. For example, in nuclear biochemical monitoring, industrial automation and environmental monitoring scenarios, a single sensor often cannot provide comprehensive information, and a centralized processing architecture is difficult to meet the requirements of real-time and high efficiency.

Existing signal processing systems suffer from significant drawbacks in several respects. First, the signal acquisition accuracy and response speed of conventional systems cannot meet the requirement of high-sensitivity monitoring, especially in complex environments of nuclear biochemistry, weak changes in the signal may be indicative of potential hazards. Second, centralized processing architectures are prone to data transmission delays and processing bottlenecks when processing large amounts of data in a distributed sensor network. In addition, the existing system has limited capability in signal fusion and fault diagnosis, and can not effectively integrate multi-source signals and perform real-time early warning and intelligent maintenance.

In order to solve the defects in the prior art, the invention provides a signal processing system, a signal processing method and a signal processing device based on a computer.

Disclosure of Invention

The invention aims to solve the problems of trivial and complex calculation flow and large calculation workload in the prior art and provides a signal processing system, a signal processing method and a signal processing device based on a computer.

In order to achieve the above object, the present invention adopts the following technical scheme that the signal processing system based on a computer comprises:

The high-precision acquisition module is used for acquiring signals through the sensor and the analog-to-digital converter;

The distributed processing module is used for realizing distributed acquisition and processing of signals through network connection and edge computing technology;

the multi-signal fusion module is used for carrying out fusion processing on the collected signals of different types through a feature extraction and fusion algorithm;

the real-time signal early warning module is used for carrying out real-time analysis on the signals through the support vector machine so as to realize real-time monitoring and early warning on abnormal signals;

The analog signal simulation module generates an analog signal through a computer simulation technology to simulate a real environment;

And the intelligent diagnosis and maintenance module monitors the running state of the system and performs fault diagnosis through the long-term and short-term memory network and provides maintenance reminding.

Further, the signal processing method based on the computer comprises the following steps:

step S1, collecting toxic agent concentration, radiation intensity, biological signals, temperature, humidity and pressure signals through a high-precision collecting module;

step S2, realizing distributed acquisition of signals through a distributed processing module, and performing preliminary data processing on edge nodes by utilizing an edge computing technology to reduce data transmission quantity;

s3, carrying out fusion processing on the collected signals of different types through a multi-signal fusion module, extracting key features and generating a comprehensive signal;

S4, analyzing the fused signals in real time through a real-time signal early warning module, detecting abnormality and sending early warning;

S5, generating an analog signal simulation real environment through an analog signal simulation module, and testing and verifying the performance of the signal processing system;

And S6, monitoring the running state of the system and diagnosing faults through the intelligent diagnosis and maintenance module, generating maintenance suggestions and reminding a user.

Further, the signal processing device based on the computer comprises:

the multi-sensor interface is used for connecting a toxin agent concentration sensor, a radiation intensity sensor, a biological signal sensor, a temperature sensor, a humidity sensor and a pressure sensor, and supports various sensor types and communication protocols;

the input interface is used for connecting the sensors and receiving signals, and supports various sensor interface standards including SPI, I2C and UART;

the network interface is used for realizing real-time transmission and distributed processing of data and supporting Ethernet and Wi-Fi communication;

The processor is used for executing the signal processing method, and is a multi-core processor and supports parallel processing;

The memory is used for storing the acquired signals, the processing results and the model, and comprises a 16GB RAM and a 512GB solid state disk;

The output interface is used for outputting processing results and early warning information, and supports various display and communication interfaces including HDMI, USB and Ethernet;

The user interface is used for displaying signal processing results, early warning information and device states, and supports touch screen operation and remote access;

The display module is used for displaying the signal processing result and the system state, and is an LCD display screen;

And the communication module is used for realizing distributed transmission and processing of signals and supporting wired and wireless communication protocols, including Ethernet, wi-Fi and 5G.

Compared with the prior art, the invention has the beneficial effects that:

The invention can realize high-precision acquisition of various signals through the high-precision acquisition module, has higher acquisition precision, shorter response time and lower error, obviously improves the quality and reliability of the signals, and can better meet the monitoring requirement in complex environments.

The invention realizes the distributed acquisition and processing of signals through the distributed processing module, not only improves the response speed of the system, but also obviously reduces the data transmission delay and processing bottleneck, can process the data in the large-scale distributed sensor network more efficiently, supports higher data transmission rate, and has stronger expansibility and flexibility.

According to the invention, through the multi-signal fusion module, different types of signals can be effectively fused, so that the accuracy and reliability of signal processing are improved, and more valuable information can be extracted.

The intelligent diagnosis and maintenance module can monitor the running state of the system in real time and timely receive maintenance reminding, thereby remarkably improving the reliability and maintenance efficiency of the system and reducing the risk of device faults and maintenance cost.

Drawings

In order to more clearly illustrate the embodiments of the invention or the technical solutions and advantages of the prior art, the following description will briefly explain the drawings used in the embodiments or the description of the prior art, and it is obvious that the drawings in the following description are only some embodiments of the invention, and other drawings can be obtained according to the drawings without inventive effort for a person skilled in the art.

FIG. 1 is a schematic diagram of a system according to an embodiment of the present invention;

FIG. 2 is a schematic flow chart of a method according to an embodiment of the present invention;

Fig. 3 is a schematic diagram of an apparatus according to an embodiment of the present invention.

Detailed Description

In order to further describe the technical means and effects of the present invention for achieving the intended purpose, the following detailed description refers to the specific embodiments, structures, features and effects of the computer-based signal processing system, method and apparatus according to the present invention with reference to the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" means that the embodiments are not necessarily the same. Furthermore, the particular features, structures, or characteristics of one or more embodiments may be combined in any suitable manner.

Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.

The following examples are for illustrative purposes only and are not intended to limit the scope of the invention.

The following describes in detail the embodiments of the computer-based signal processing system, method and apparatus provided by the present invention with reference to the accompanying drawings.

Examples

Referring to fig. 1, a schematic diagram of a system configuration provided in an embodiment of the present invention, which provides a signal processing system based on a computer in the embodiment of the present invention, includes:

1. High-precision acquisition module

The high-precision acquisition module comprises a signal collection unit, an analog-to-digital conversion unit and a signal conditioning unit;

The signal collection unit comprises a toxin agent concentration sensor, a radiation intensity sensor, a biological signal sensor, a temperature sensor, a humidity sensor and a pressure sensor, and is used for realizing signal collection, wherein the response time of the signal collection is not more than 1 second, and the error is lower than 1%;

the analog-to-digital conversion unit comprises an analog-to-digital converter, which is used for converting the acquired analog signals into digital signals, and has 16-bit resolution and a sampling rate of 100kHz;

the signal conditioning unit comprises a filter circuit and an amplifying circuit, wherein the filter circuit is used for filtering, amplifying and normalizing acquired signals, the filter circuit adopts a low-pass filter, the cut-off frequency is 10Hz, and the gain range of the amplifying circuit is 1-100 times.

The specific indexes of the sensitivity and the precision of the sensor are as follows:

The sensitivity of the toxin agent concentration sensor is that the variation of the output signal is not lower than 10mV/ppm, the detection limit is 0.01ppm, the response time is not more than 1 second, and the measurement accuracy is +/-0.1%.

The radiation intensity sensor has sensitivity that the variation quantity of output signals is not lower than 5mV/mSv/h, the detection limit is 0.01mSv/h, the response time is not more than 1 second, and the measurement accuracy is +/-0.5%.

The sensitivity of the biological signal sensor is that the variation amount of the output signal is not lower than 1 mV/mu V per mu V, the detection limit is 0.1 mu V, the response time is not more than 1 second, and the measurement accuracy is +/-0.2%.

The sensitivity of the temperature sensor is that the variation of the output signal is not lower than 10mV/°C, the detection limit is 0.1 ℃, the response time is not more than 1 second, and the measurement accuracy is +/-0.2 ℃.

The sensitivity of the humidity sensor is that the variation of the output signal is not lower than 5mV/%, the detection limit is 0.1%, the response time is not more than 1 second, and the measurement accuracy is +/-1%.

The sensitivity of the pressure sensor is that the variation of output signals is not lower than 10mV/kPa per kPa, the detection limit is 0.1kPa, the response time is not more than 1 second, and the measurement accuracy is +/-0.5 kPa.

16-Bit resolution means that the analog-to-digital converter can convert the analog signal into 65536 (2-16) different digital values, which can capture small changes in the signal.

Sampling rate of 100kHz the sampling rate of 100kHz means that the analog-to-digital converter can sample 100000 times per second, so that the analog-to-digital converter can capture 50kHz signals, cover the frequency range of signals common in most industrial and environmental monitoring, and ensure that rapid changes of signals are captured in time, thereby realizing rapid response.

The cut-off frequency of the low-pass filter is 10Hz, the low-pass filter has the main function of removing high-frequency noise, the high-frequency noise can be effectively removed by setting the cut-off frequency of 10Hz, the main characteristics of the signals are reserved, and the high-precision processing of the signals is ensured

The gain range of the amplifying circuit is 1-100 times, for weak signals, high gain is needed to enhance the signal strength, the gain range of 1-100 times can meet the requirement, the signals are ensured not to lose important information due to over-weak in subsequent processing, for strong signals, low gain is needed to avoid signal saturation, the gain range of 1-100 times can be flexibly adjusted, and the signals are ensured to be in a proper range.

2. Distributed processing module

The distributed processing module comprises a network connection unit, a data transmission unit and a distributed computing unit;

The network connection unit is used for connecting the sensor and the real-time signal processing unit through a wired network and a wireless network to construct a distributed signal processing network, wherein the wired network comprises Ethernet, optical fiber communication, an industrial bus and serial communication, and the wireless network comprises Wi-Fi, 5G, zigBee, bluetooth, loRa and NB-IoT;

The data transmission unit is used for realizing signal transmission and supporting the functions of data compression, data encryption and data verification, wherein the transmission speed is higher than 100Mbps, the data compression comprises Huffman coding, wavelet transformation and discrete cosine transformation DCT, the data encryption comprises advanced encryption standard AES, an asymmetric encryption algorithm RSA and an elliptic curve encryption algorithm ECC, and the data verification comprises cyclic redundancy check CRC, CRC32 and parity check;

And the distributed computing unit is used for realizing the distributed processing of the signals, the distributed processing comprises preliminary data processing at edge nodes by an edge computing technology, the data transmission quantity is reduced, the processed result is uploaded, and the preliminary data processing comprises data preprocessing, feature extraction and simple analysis.

It should be noted that, the wired network refers to a network for data transmission through physical media (cables and optical fibers), and the physical media provide stable connection and higher data transmission rate, and have the characteristics of high stability, fast transmission rate and strong security, and are applied to scenes of Local Area Networks (LANs), wide Area Networks (WANs) and data centers, which need high data transmission rate and stability.

The wireless network refers to a network for carrying out data transmission through radio waves, does not need a physical medium to connect devices, has the characteristics of high flexibility, simple deployment and low cost, and is applied to the fields of mobile communication, internet of things, intelligent home and remote office.

The classification of wireless networks includes Wireless Wide Area Networks (WWANs) that can be widely covered across cities and even countries, wireless Local Area Networks (WLANs) that are local areas for homes, offices or schools, wireless Personal Area Networks (WPANs) that are small in coverage, typically within 10 meters, wireless Metropolitan Area Networks (WMANs) that are in between WLANs and WWANs for urban or large campuses, and Wireless Body Area Networks (WBANs) that are used to connect medical monitoring devices attached to the human body.

The data transmission rate is an important index for measuring the data transmission speed, and the data transmission rate is expressed in units of Mbps (megabits per second), and is a key parameter for evaluating the network performance, and a higher data transmission rate means that more data can be transmitted in the same time, so that the overall transmission efficiency is improved.

The data transmission rate higher than 100Mbps to 100Mbps can meet the requirements of most scenes, is a reference line of high-efficiency transmission, can support simultaneous online of multiple devices and meets the requirements of large-scale device connection.

Data compression is a technique that reduces the amount of data by converting the original data into a more compact form through a specific algorithm while preserving the original information or characteristics of the data.

Huffman coding is a lossless data compression algorithm, an optimal binary tree (Huffman tree) is constructed according to the occurrence frequency of characters in data, characters with high occurrence frequency are coded by using the optimal binary tree, characters with low occurrence frequency are coded by using shorter codes, and characters with low occurrence frequency are coded by using longer codes, so that the whole size of the data can be effectively reduced.

Wavelet transformation, which is a mathematical tool that analyzes and compresses signals by decomposing them into wavelets of different frequency components, is suitable for compression of image and audio signals, and is capable of capturing local features of the signals and preserving those features during compression.

The DCT is a mathematical transformation for converting signals from a spatial domain to a frequency domain, is suitable for image and video compression, and can remove redundant information in images and retain main characteristics of the images.

Data encryption is a method of protecting data from unauthorized access by converting the original data into ciphertext that cannot be directly understood.

The advanced encryption standard AES, a symmetric encryption algorithm, uses the same key to encrypt and decrypt, has high efficiency and security, and is widely applied to various encryption scenes.

The asymmetric encryption algorithm RSA is an asymmetric encryption algorithm based on a large number of decomposition problems, and uses a pair of keys (a public key and a private key) for encryption and decryption, wherein the public key is used for encrypting data, and the private key is used for decrypting data.

Elliptic Curve Cryptography (ECC) is an asymmetric cryptography based on elliptic curve math, and can provide security comparable to RSA by using a smaller key length.

Data verification is a method of ensuring data integrity and accuracy by adding a check code during data transmission or storage to detect data errors.

The cyclic redundancy check CRC is a check method based on polynomial division, and a receiver can use the same polynomial to verify the integrity of data by regarding the data as a polynomial and calculating the rest of numbers to generate a check code.

CRC32 is a variation of the CRC algorithm, using a 32-bit check code to detect data errors.

Parity, a simple data verification method, detects errors by checking whether the number of 1 s in the data is odd or even.

The edge computing technology is a distributed computing architecture, and moves data processing, storage and service functions to edge positions of data generation instead of completely relying on a centralized data center or cloud computing platform, wherein the edge positions refer to any place or source close to data generation, including mobile equipment, internet of things equipment, a gateway and a small data center;

The method comprises the steps of preliminary data processing, feature extraction and simple analysis, wherein the data preprocessing mainly comprises the steps of cleaning, denoising and formatting the original data to improve the quality and usability of the data, the feature extraction is used for extracting useful features and information from the data so as to facilitate subsequent analysis and processing, and the simple analysis is used for carrying out preliminary analysis and judgment on the data to find rules and anomalies of the data.

3. Multi-signal fusion module

The multi-signal fusion module comprises a feature extraction unit, a fusion processing unit and a fusion algorithm selection unit;

The characteristic extraction unit is used for extracting key characteristics from the acquired signals by adopting wavelet transformation and Fourier transformation algorithms, wherein the key characteristics comprise frequency characteristics, time domain characteristics and waveform characteristics;

the fusion processing unit is used for carrying out fusion processing on the extracted features through a fusion algorithm to realize signal fusion, wherein the fusion algorithm comprises weighted average, bayesian fusion and convolutional neural network;

And a fusion algorithm selection unit for selecting a fusion algorithm based on signal types including a toxin agent concentration signal, a radiation intensity signal, a biological signal, a temperature signal, a humidity signal, and a pressure signal.

The wavelet transformation is a mathematical tool capable of decomposing a signal into components with different frequencies and scales, and frequency characteristics can be extracted from the signal through wavelet transformation, so that the energy distribution of the signal on different frequencies can be reflected.

The Fourier transform is a mathematical tool, which can convert signals from time domain to frequency domain, and can obtain spectrograms of the signals through Fourier transform, thereby extracting frequency characteristics, and can be used for extracting time domain characteristics including mean value, variance and peak value of the signals.

The key features include frequency features reflecting the characteristic of the signal in the frequency domain, time domain features describing the time variation law of the signal and waveform features showing the shape and structure of the signal.

The weighted average is a simple fusion algorithm, and the weighted summation is carried out on the characteristics according to the weight of each signal, so that the weighted average is suitable for the condition of stronger correlation between signals.

Bayesian fusion, namely based on the bayesian theorem, the fusion is realized by calculating the posterior probability of each signal, the uncertainty between the signals can be processed, and the accuracy of the fusion result is improved.

The convolutional neural network is a deep learning algorithm, can extract higher-level signals from the features through learning, and can be used for extracting and fusing complex feature modes in signal fusion.

The concentration signal of the chemical toxic agent has a definite threshold value and has high real-time requirement, and the fusion algorithm selects a Bayesian fusion algorithm, is suitable for processing data with uncertainty, and can update posterior probability according to priori knowledge and observation data, so that accurate toxic agent concentration estimation is provided.

The method has the advantages that the method is simple and easy to realize, and is suitable for the condition that the change of the radiation intensity signal is stable, and the convolutional neural network can process more complex radiation intensity change modes and improve the accuracy of fusion results.

The biological signals comprise heart rate, electroencephalogram and electromyogram, have the characteristics of nonlinearity, non-stationarity and high noise, and the fusion algorithm selects a convolutional neural network, so that complex features in the biological signals can be automatically extracted, noise and interference in the signals are processed, and the accuracy and the robustness of fusion results are improved.

The temperature signal is an environmental or object temperature measurement signal, has continuity and stability, is influenced by environmental noise, and is suitable for the condition that the temperature signal changes more stably and has smaller noise by selecting a weighted average algorithm by a fusion algorithm.

The humidity signal is an environment humidity measurement signal, has continuity and stability, is influenced by environment noise, and the fusion algorithm selects a weighted average algorithm to carry out fusion processing, so that noise and interference in the humidity signal can be processed, and the accuracy of a fusion result is improved.

The pressure signal, namely an atmospheric pressure or liquid pressure measurement signal, has real-time performance and dynamic performance, and the fusion algorithm selects a convolutional neural network, can process more complex pressure change modes and is suitable for different working environments and conditions.

4. Real-time signal early warning module

The real-time signal early warning module comprises a real-time signal processing unit, an intelligent analysis unit and an early warning decision unit;

The real-time signal processing unit is used for processing the acquired signals in real time, and accelerating the signal processing process by adopting the FPGA to realize the signal processing with the delay lower than 10 milliseconds;

The intelligent analysis unit is used for carrying out real-time analysis and anomaly detection on the signals by adopting a support vector machine, wherein the real-time analysis comprises pattern recognition, trend analysis and signal classification, and the anomaly detection comprises pattern anomaly detection, trend anomaly detection and signal anomaly recognition;

And the early warning decision unit judges whether an early warning signal is sent out according to the analysis result, wherein the early warning signal comprises an alarm type and alarm content, the alarm type comprises an audio alarm, a visual alarm and a vibration alarm, and the alarm content comprises an abnormal type, an abnormal position, an abnormal severity and recommended measures.

It should be noted that FPGA acceleration, a Field Programmable Gate Array (FPGA), is a hardware accelerator that allows multiple tasks to be processed in parallel, thereby significantly increasing the speed of signal processing.

The processing delay is lower than 10 milliseconds, namely, the processing delay is lower than 10 milliseconds, so that the system can process the input signal very rapidly and generate output, the time difference between signal acquisition and early warning signal emission can be reduced by the shorter delay time, the risk of signal loss or delay caused by the processing delay is reduced, the capturing and processing capacity of the system to the signal is improved, and the reliability of the system is enhanced.

Support Vector Machine (SVM) A support vector machine is a machine learning algorithm, particularly suitable for classification and regression problems, and in an intelligent analysis unit, the SVM is used for real-time analysis and anomaly detection of signals, can identify patterns, trends and categories in the signals, and judges whether the signals are normal or not according to the characteristics.

Pattern recognition, which refers to a process of recognizing a specific pattern or characteristic from an input signal, can be used to recognize a normal operation state of a device, an abnormal signal pattern.

Trend analysis refers to the process of predicting future trend of an input signal by carrying out time sequence analysis on the input signal, and is helpful for timely finding potential problems in the signal, including device performance reduction and fault early warning.

Signal classification, which refers to the process of classifying input signals into different categories or types, may be used to distinguish between different types of signals, including normal signals, abnormal signals, noise signals.

And the abnormal mode detection is a process of identifying signals which deviate from the normal mode obviously, and abnormal signals which do not accord with the normal mode can be timely found.

Trend anomaly detection, namely a process of identifying signals which deviate from expected trends obviously, and can predict future trend of the signals and discover potential anomalies in time.

Signal abnormality recognition, namely further analyzing abnormal signals, wherein the process of determining the abnormality type and the cause of the abnormal signals can provide information for subsequent fault investigation and maintenance decision.

Audible alert-refers to the process of alerting the user to the attention by emitting a specific audible signal.

Visual alert-refers to the process of alerting the user to the attention by displaying a specific visual signal (indicator light, screen display).

Vibration alert-refers to the process of alerting the user to the attention by emitting a vibration signal.

Abnormality type, which is to say, the result of classifying the abnormal signal, including device fault, intrusion detection and performance degradation, and knowing the abnormality type is helpful for users to quickly judge the nature and severity of the problem.

Abnormal position refers to a specific position or region where an abnormal signal occurs, which is helpful for a user to quickly locate the problem and take corresponding treatment measures.

Abnormal severity refers to the influence degree of an abnormal signal on a system, and knowing the abnormal severity is helpful for users to judge the urgency and priority of the problem, so that resources are reasonably allocated for processing.

The proposal measures are that processing proposal or solution provided for the user according to the abnormal type, position and severity information can help the user to quickly solve the problem and restore the normal operation of the system.

5. Analog signal simulation module

The analog signal simulation module comprises a scene generation unit, a signal generation unit and a simulation control unit;

The scene generation unit is used for generating a scene of the analog signal according to preset conditions, wherein the preset conditions comprise environment parameters, event parameters, analog requirements and technical parameters;

the signal generating unit is used for generating a simulation signal according to the generated scene, simulating signal characteristics in a real environment through a signal generating algorithm, wherein the signal generating algorithm comprises a simulation algorithm based on a physical model, a generation algorithm based on data driving, a random signal generating algorithm and a parameterized signal generating algorithm;

and the simulation control unit is used for controlling the generation process of the analog signal and supporting adjustment of signal parameters, wherein the signal parameters comprise frequency, amplitude, phase, time delay, pulse width, duty ratio, noise level, harmonic distortion, modulation index and signal bandwidth.

The environmental parameters include physical environment such as temperature, humidity, air pressure and wind speed, geographical environment such as topography, topography and building distribution, and meteorological conditions such as rainfall, haze and visibility.

Event parameters include event type, toxic agent leakage, radiation leakage, bio-threat, fire, natural disaster, industrial accident, event intensity, leakage rate, radiation intensity, virus concentration, fire intensity, seismic intensity, event location, specific location (coordinates, area) where the leakage source or event occurred, event time, specific time or time frame of the event occurrence.

The simulation requirements comprise the purposes of training, testing, evaluating and researching, the simulation accuracy of high fidelity, medium fidelity and low fidelity, the simulation range of local area, whole field and city range and the user requirements of customization requirements of specific users or organizations.

The technical parameters comprise sensor types, namely a toxin agent concentration sensor, a radiation intensity sensor, a biological signal sensor, a temperature sensor, a humidity sensor and a pressure sensor, and data formats, namely a data format and a transmission protocol of analog signals.

Simulation algorithm based on physical model to simulate signal generating and propagation process based on physical law and principle, and is suitable for use in the scene needing accurate simulation of signal characteristic.

The data-driven generation algorithm is used for generating the analog signal by utilizing the existing data, generating the analog signal similar to the real signal by analyzing the statistical characteristics and rules of the data, and is suitable for scenes with rich data and easy acquisition.

And a random signal generation algorithm is used for generating random signals according to probability distribution, and is suitable for signals with strong randomness of analog noise and interference.

And the parameterized signal generation algorithm allows a user to adjust parameters of the signal according to requirements, provides a flexible signal generation mode, and is suitable for scenes needing to customize signal characteristics.

The frequency refers to the number of vibrations per second of the signal, which determines the periodicity of the signal.

Amplitude refers to the peak or maximum value of a signal, reflecting the strength of the signal.

The phase refers to the position of the waveform of the signal relative to a certain reference point, and determines the phase relation of the signal.

Time delay refers to the time required for a signal to travel from generation to a point of reception.

Pulse width, which refers to the duration of a pulse signal, is an important feature for a pulse signal.

Duty cycle, which refers to the proportion of the high level in the pulse signal, has an important effect on the characteristics and application of the pulse signal.

Noise level, which refers to the intensity of random noise mixed in a signal, has an effect on the sharpness and accuracy of the signal.

Harmonic distortion refers to the ratio of harmonic components to fundamental components in a signal, reflecting the degree of distortion of the signal.

Modulation index refers to the ratio of the amplitude of a modulated signal to the carrier signal.

The signal bandwidth refers to the frequency range occupied by the signal, and determines the transmission capacity and the anti-interference capacity of the signal.

6. Intelligent diagnosis and maintenance module

The intelligent diagnosis and maintenance module comprises a state monitoring unit, a fault diagnosis unit and a maintenance reminding unit;

The state monitoring unit is used for monitoring the running state of the system in real time, evaluating the health condition of the system in real time through the parameters of the acquisition device during running, and transmitting the monitoring data to the fault diagnosis unit in real time for analysis, wherein the parameters comprise the temperature, the current, the voltage, the running time and the sensor state of the device;

The fault diagnosis unit is used for modeling the running state of the device through the long-term and short-term memory network, identifying an abnormal mode in the running process of the device, and realizing fault diagnosis, wherein the fault diagnosis comprises the steps of judging the fault type, the fault position and the severity;

and the maintenance reminding unit is used for generating maintenance advice according to the diagnosis result, displaying the health state of the device and the maintenance advice through the user interface, wherein the maintenance advice comprises maintenance steps, required tools and notes.

It should be noted that, the state monitoring unit collects the operation parameters through the sensor network deployed on the device, and the sensor includes a temperature sensor, a current sensor and a voltage sensor, so that the operation state data of the device can be obtained in real time, and in order to ensure real-time performance, the state monitoring unit collects 2 times of data per second, so as to ensure that the tiny change of the state of the device can be captured in time.

The long-term memory network LSTM is a special type of circulating neural network, is specially designed for solving the problems of gradient disappearance and gradient explosion of the traditional RNN when processing long-sequence data, and can effectively capture long-term dependency relationship in time-sequence data by introducing a gate mechanism.

LSTM modeling, namely inputting the extracted characteristic sequence into an LSTM network, and building a model of the running state of the device by learning a mode in time sequence data, wherein the method comprises the following specific steps of:

1. And (3) inputting a sequence, namely taking time sequence data as an input and inputting the time sequence data into an LSTM network.

2. And updating the hidden state, namely updating the hidden state by the LSTM through the synergistic effect of the input door, the forget door and the output door.

3. And outputting prediction, namely judging whether an abnormality exists by an output layer of the LSTM according to the running state of the hidden state prediction device.

Judging the fault type comprises the following steps:

overheat fault if the temperature change rate exceeds a set threshold (5 ℃ rise per second), or the temperature peak exceeds a normal range (exceeding 55 ℃), then it is determined as overheat fault.

Cooling system failure if the temperature fluctuation range increases abnormally (exceeds 20 ℃) and the temperature continues to rise, the cooling system failure.

Short-circuit fault if the current change rate increases sharply (2A per second), or the current peak exceeds the normal range (exceeds 3A), it is determined as a short-circuit fault.

Power failure if the current fluctuation range increases abnormally (exceeds 1A) and the current continues to drop, the power failure.

An excessive voltage fault is determined if the rate of change of the voltage increases sharply (10V per second rise), or the voltage peak exceeds the normal range (exceeding 260V).

An excessively low voltage failure is determined if the voltage change rate drops sharply (10V per second), or the voltage peak is below the normal range (below 200V).

Sensor failure if the sensor state becomes "off-line" or the returned data anomaly rate exceeds a set threshold (exceeds 10%), then a sensor failure is determined.

Judging the fault position comprises the following steps:

local overheat, in which if a temperature sensor at a specific position detects an abnormally high temperature and surrounding sensors do not detect a similar abnormality, the fault location is determined to be in the vicinity of the sensor.

And (3) overall overheat, namely judging that the fault is likely to be in the overall heat dissipation system of the device if the temperature sensors at a plurality of positions detect abnormally high temperatures.

Local short-circuiting if a current sensor at a specific position detects an abnormally high current and surrounding sensors do not detect a similar abnormality, the fault position is judged to be in the vicinity of the sensor.

And (3) overall power failure, namely judging that the failure is likely to be in an overall power system of the device if the current sensors at a plurality of positions detect abnormally low currents.

The local voltage is too high, and if the voltage sensor at a specific position detects an abnormally high voltage and the surrounding sensors do not detect similar abnormality, the fault position is judged to be near the sensor.

The overall voltage is too low, and if the voltage sensors at a plurality of positions detect abnormally low voltages, the fault is judged to be possible in the overall power supply system of the device.

Local sensor failure if the sensor state of a specific position is changed to be off-line, the position of the failure is judged to be in the sensor.

And (3) systematic sensor faults, namely judging that the faults are likely to occur in a sensor network or a data transmission system of the device if the sensors at a plurality of positions are abnormal in state.

Judging the severity comprises the following steps:

The severity of the temperature anomaly includes a minor fault that the magnitude of the temperature anomaly is small (5 ℃ above the normal range) and the duration is short (less than 1 minute), a moderate fault that the magnitude of the temperature anomaly is medium (10 ℃ above the normal range) and the duration is long (more than 5 minutes), and a severe fault that the magnitude of the temperature anomaly is large (20 ℃ above the normal range) and the duration is long (more than 10 minutes).

Current anomaly severity includes minor faults with small current anomaly magnitude (0.5A above normal range) and short duration (less than 1 minute), moderate faults with medium current anomaly magnitude (1A above normal range) and long duration (5 minutes), severe faults with large current anomaly magnitude (2A above normal range) and long duration (10 minutes);

voltage anomaly severity includes minor faults with small voltage anomaly magnitude (5V above normal range) and short duration (less than 1 minute), moderate faults with medium voltage anomaly magnitude (10V above normal range) and long duration (more than 5 minutes), severe faults with large voltage anomaly magnitude (20V above normal range) and long duration (more than 10 minutes).

The severity of sensor state anomalies includes minor faults, moderate sensor state anomalies (5% -10%) and a large number of affected sensors (3-5), severe faults, high sensor state anomalies (more than 10%) and a large number of affected sensors (more than 5).

The health status of the device is important information displayed by the maintenance reminding unit and is used for intuitively informing the user of the current running condition of the device, and the health status comprises the following aspects:

the temperature conditions include normal device temperature in the normal range (30 ℃ to 50 ℃), warning that device temperature is near the upper limit of the normal range (near 55 ℃) and fault that device temperature exceeds the normal range (exceeding 60 ℃).

The current state includes normal device current within a normal range (1A to 2A), warning that device current is near the upper limit of the normal range (near 2.5A), fault that device current exceeds the normal range (exceeds 3A).

The voltage conditions include normal device voltage within a normal range (220V to 240V), warning that the device voltage is near the upper or lower limit of the normal range (near 210V or 250V), fault that the device voltage exceeds the normal range (below 200V or above 260V).

The runtime status includes normal device operation time within an expected range (8 hours per day), warning that device operation time is approaching an expected upper limit (over 10 hours), failure that device operation time exceeds an expected upper limit (over 12 hours).

Sensor states include normal, all sensors on-line and returning valid data, warning that part of the sensor state is abnormal (data anomaly rate exceeds 5%), failure, sensor off-line or data anomaly rate exceeds 10%.

The maintenance reminding unit can also provide the points to be paid attention in the maintenance process so as to ensure the safety and the effectiveness of the maintenance. The following are common notes:

the maintenance steps include the following:

Overheat fault, checking the cooling system of the device to work normally, cleaning the radiator or the cooling fan of the device, and checking the surrounding of the device to have obstacle for heat dissipation.

Short-circuit fault, checking whether the power line and the connection point of the device are damaged, checking whether the short-circuit phenomenon exists in the device, and replacing the damaged fuse or the circuit protection device.

The over-high or over-low voltage faults are that whether the power supply of the device is normal or not is checked, whether the power line is damaged or in poor contact is checked, and the power supply voltage of the device is adjusted to be in a normal range.

The sensor faults are that whether the sensor is on line or not is checked, the sensor is reconnected, whether the sensor is damaged or not is checked, the sensor is replaced if necessary, and the sensor is calibrated to ensure that the returned data is accurate.

The running time is excessively long, namely, whether the device runs continuously for a long time and intermittently runs if necessary, whether the device needs maintenance or not, and whether the running time is abnormally prolonged due to the fact that the device has faults or not is checked.

The required tools include the following:

the universal tools comprise a screwdriver (cross and straight line), a wrench, a universal meter and pliers.

The electric tool comprises a voltmeter, an ammeter, an insulating tape and an electric wire stripper.

And the cooling system tool comprises a cleaning brush, a hot air gun and cooling liquid.

Sensor tool, calibration device, sensor tester, sensor installation tool.

Other tools, antistatic wrist rings, protective gloves and safety glasses.

The notes include the following:

Safety precautions, ensuring that the device has been powered off before any maintenance operations are performed, wearing appropriate protective equipment such as gloves and safety glasses, ensuring that the working area is dry, avoiding the risk of electric shock.

Operational precautions are ensuring that components are removed or installed in the correct order, ensuring that new components are compatible with the original components when the components are replaced, and avoiding the use of excessively humid cleaning tools when cleaning the device so as not to damage the electronic components.

Environmental precautions, ensuring that there is sufficient space around the device for maintenance operations, ensuring that there are no inflammable items around the device, avoiding fire risks, cleaning the working area after maintenance is complete, ensuring that there are no tools or impurities left behind.

For the computer-based signal processing method in the present embodiment, the method includes the steps of:

and S1, collecting toxic agent concentration, radiation intensity, biological signals, temperature, humidity and pressure signals through a high-precision collecting module.

The toxic agent concentration signal is collected by a toxic agent concentration sensor with high sensitivity and is used for detecting whether toxic gas or chemical substances exist in the environment.

The radiation intensity signal is collected by a radiation intensity sensor, the radiation level in the environment is monitored, and the method is suitable for nuclear facilities and medical radiology scenes.

Biological signals, such as heart rate, respiratory rate and the like, are collected, and the biological signal acquisition device is suitable for medical health monitoring and biological experimental scenes.

And the temperature signal is used for acquiring temperature data of the environment or the device through a temperature sensor and is used for monitoring the environment and the running state of the industrial device.

The humidity signal is obtained by using a humidity sensor, and is suitable for meteorological monitoring and storage environment control scenes.

Pressure signal, which is to collect pressure data by pressure sensor for industrial pipeline pressure monitoring and meteorological air pressure monitoring

And S2, realizing distributed acquisition of signals through a distributed processing module, and performing preliminary data processing on edge nodes by utilizing an edge computing technology to reduce data transmission quantity.

The network connection is to connect a plurality of sensor nodes through wired (Ethernet, optical fiber) and wireless (Wi-Fi, 5G) networks to construct a distributed signal acquisition network.

And (3) edge calculation, namely performing preliminary data processing on edge nodes close to the sensor, reducing the data transmission quantity and improving the system efficiency.

The primary data processing comprises data compression, data filtering and data preprocessing, wherein the data compression compresses data by adopting Huffman coding and wavelet transformation algorithm to reduce storage and transmission cost, the data filtering removes high-frequency noise through a low-pass filter and retains main characteristics of signals, and the data preprocessing comprises normalization processing of acquired signals to enable the acquired signals to be comparable in subsequent processing.

And S3, carrying out fusion processing on the collected signals of different types through a multi-signal fusion module, extracting key features and generating a comprehensive signal.

The feature extraction is to extract key features including frequency features, time domain features and waveform features from different types of signals by using wavelet transform and Fourier transform algorithm.

And the fusion algorithm is to perform fusion processing on the extracted features by adopting weighted average, bayesian fusion and convolutional neural network algorithm to generate a comprehensive signal.

The integrated signal can reflect the state of the environment or the device more comprehensively, and the characteristics of the signal can be enhanced and the analyzability of the signal can be improved through the integration processing.

And S4, analyzing the fused signals in real time through a real-time signal early warning module, detecting abnormality and sending early warning.

The early warning mechanism comprises early warning signals, wherein the early warning signals comprise sound alarms, visual alarms and vibration alarms, the early warning signals comprise abnormal types, abnormal positions, abnormal severity and recommended measures, and the early warning signals are judged according to analysis results so as to help a user to respond quickly.

And S5, generating an analog signal simulation real environment through an analog signal simulation module for testing and verifying the performance of the signal processing system.

It should be noted that the purpose of the analog signal simulation is to test the performance of the system by generating an analog signal to simulate the real environment, and to verify the reliability of the system by ensuring that the system can operate stably in various complex environments.

The fault diagnosis method includes modeling the running state of the device by using the long-short-period memory network, identifying abnormal modes, and judging the fault type, the fault position and the severity.

The prompting mode includes displaying the health state of the device and the maintenance suggestion through a user interface, and prompting the user in a mode of supporting short messages, mails and the like.

For the computer-based signal processing apparatus in the present embodiment, it includes:

the multi-sensor interface is used for connecting a toxin agent concentration sensor, a radiation intensity sensor, a biological signal sensor, a temperature sensor, a humidity sensor and a pressure sensor, and supports various sensor types and communication protocols.

It should be noted that the design reasons of the multi-sensor interface include versatility, that is, the multi-sensor interface can be connected with various sensors to realize multi-parameter monitoring, and is suitable for complex environments, and compatibility, that is, support various communication protocols, that is, ensure compatibility with sensors of different manufacturers and types, the communication protocols include SPI and I2C, UART.

The input interface is used for connecting the sensors and receiving signals, and supports various sensor interface standards including SPI, I2C and UART.

It should be noted that the input interface design reasons include versatility, supporting multiple interface standards, ensuring that the device can be connected with various sensors, and flexibility, being capable of receiving different types of signals and being suitable for multiple application scenarios.

And the network interface is used for realizing real-time transmission and distributed processing of data and supporting Ethernet and Wi-Fi communication.

It should be noted that, the network interface design reasons include high-efficiency transmission, supporting high-speed communication protocol, ensuring real-time and reliability of data transmission, and distributed processing, supporting distributed processing architecture, and improving processing capacity and efficiency of the system.

it should be noted that the processor design reasons include high performance, that is, the multi-core processor can process a large amount of data to meet the requirement of complex signal processing, and parallel processing, that is, support parallel processing to improve the processing speed and efficiency of the system.

The memory is used for storing the acquired signals, the processing results and the model, and comprises a 16GB RAM and a 512GB solid state disk.

It should be noted that the reasons for designing the memory include large-capacity storage, meeting the requirement of large-scale data storage, ensuring the integrity and availability of data, and high-efficiency reading and writing, namely, the solid state disk has the capability of fast reading and writing, thus improving the overall performance of the system.

The output interface is used for outputting processing results and early warning information, and the output interface supports various display and communication interfaces including HDMI, USB and Ethernet.

It should be noted that the design reasons of the output interface include versatility, supporting multiple interfaces, ensuring that the device can be connected with various devices, and flexibility, being able to meet the needs of different users and providing multiple output modes.

And the user interface is used for displaying signal processing results, early warning information and device states and supporting touch screen operation and remote access.

It should be noted that the user interface design reasons include user friendliness, namely, providing an intuitive user interface, facilitating user operation and monitoring, and remote monitoring, namely, supporting remote access, and improving flexibility and practicability of the system.

The display module is used for displaying the signal processing result and the system state, and the display module is a high-resolution LCD display screen.

The display module design reasons include clear display, wherein the high-resolution display screen can clearly display complex processing results and system states, and the high-quality visual effect is that a high-quality display technology is supported, so that user experience is improved.

It should be noted that, the reasons for designing the communication module include high-efficiency transmission, supporting multiple communication protocols, ensuring high-efficiency and reliability of data transmission, and distributed processing, supporting distributed processing, and improving flexibility and expandability of the system.

The foregoing embodiments are merely illustrative of the technical solutions of the present application, and not restrictive, and although the present application has been described in detail with reference to the foregoing embodiments, it should be understood by those skilled in the art that modifications may still be made to the technical solutions described in the foregoing embodiments or equivalent substitutions of some technical features thereof, and that such modifications or substitutions do not depart from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A computer-based signal processing system, comprising:

2. The computer-based signal processing system of claim 1, wherein the high-precision acquisition module comprises:

The analog-to-digital conversion unit comprises an analog-to-digital converter, and is used for converting the acquired analog signals into digital signals, wherein the analog-to-digital converter has 16-bit resolution and the sampling rate is 100kHz;

3. The computer-based signal processing system of claim 1, wherein the distributed processing module comprises:

The distributed computing unit is used for realizing the distributed processing of the signals, the distributed processing comprises preliminary data processing at edge nodes by an edge computing technology, the data transmission quantity is reduced, and the processed result is uploaded, wherein the preliminary data processing comprises data preprocessing, feature extraction and simple analysis.

4. The computer-based signal processing system of claim 1, wherein the multi-signal fusion module comprises:

And a fusion algorithm selection unit that selects a fusion algorithm based on signal types including a toxic agent concentration signal, a radiation intensity signal, a biological signal, a temperature signal, a humidity signal, and a pressure signal.

5. The computer-based signal processing system of claim 1, wherein the real-time signal pre-warning module comprises:

And the early warning decision unit is used for judging whether an early warning signal is sent out according to the analysis result, wherein the early warning signal comprises an alarm type and alarm content, the alarm type comprises an audio alarm, a visual alarm and a vibration alarm, and the alarm content comprises an abnormal type, an abnormal position, an abnormal severity and recommended measures.

6. The computer-based signal processing system of claim 1, wherein the analog signal emulation module comprises:

The scene generation unit is used for generating scenes of the analog signals according to preset conditions, wherein the preset conditions comprise environment parameters, event parameters, analog requirements and technical parameters;

a signal generating unit for generating a simulation signal according to the generated scene, and simulating signal characteristics in a real environment through a signal generating algorithm, wherein the signal generating algorithm comprises a simulation algorithm based on a physical model, a generation algorithm based on data driving, a random signal generating algorithm and a parameterized signal generating algorithm;

And the simulation control unit is used for controlling the generation process of the analog signal and supporting adjustment of signal parameters, wherein the signal parameters comprise frequency, amplitude, phase, time delay, pulse width, duty cycle, noise level, harmonic distortion, modulation index and signal bandwidth.

7. The computer-based signal processing system of claim 1, wherein the intelligent diagnostic and maintenance module comprises:

The state monitoring unit is used for monitoring the running state of the system in real time, evaluating the health condition of the system in real time through the parameters of the acquisition device during running, and transmitting the monitoring data to the fault diagnosis unit in real time for analysis, wherein the parameters comprise the temperature, current, voltage, running time and sensor state of the device;

The fault diagnosis unit is used for modeling the running state of the device through the long-term and short-term memory network, identifying an abnormal mode in the running process of the device and realizing fault diagnosis, wherein the fault diagnosis comprises the steps of judging the fault type, the fault position and the severity;

and the maintenance reminding unit is used for generating maintenance advice according to the diagnosis result, and displaying the health state of the device and the maintenance advice through a user interface, wherein the maintenance advice comprises maintenance steps, required tools and notes.

8. A computer-based signal processing method, comprising the steps of:

9. A computer-based signal processing apparatus, comprising: