WO2022100084A1

WO2022100084A1 - Apparatus for detecting abnormal vibrations in industrial device

Info

Publication number: WO2022100084A1
Application number: PCT/CN2021/099699
Authority: WO
Inventors: 孙广宇; 罗国杰; 韩平; 孙康睿; 张波; 李亚军; 曹玉龙; 李加敏
Original assignee: 杭州未名信科科技有限公司; 浙江省北大信息技术高等研究院
Priority date: 2020-11-11
Filing date: 2021-06-11
Publication date: 2022-05-19
Also published as: CN112697267A

Abstract

Disclosed is an apparatus for detecting abnormal vibrations in an industrial device. The apparatus comprises an industrial device, an IoT node module, and an intelligent edge computing platform. The industrial device is connected to the IoT node module, and the intelligent edge computing platform is in communication connection with the IoT node module. The IoT node module performs detection on vibration data of the industrial device, performs analysis, and determines whether or not vibrations in the industrial device are abnormal. The IoT node module comprises an embedded soft core. The embedded soft core has a built-in neural network and a built-in self-learning algorithm. The self-learning algorithm is used to achieve unsupervised online training of the neural network. In the present invention, the embedded soft core has the built-in neural network and the built-in self-learning algorithm, so that operational data can be collected online and used to perform unsupervised online training of the neural network, thereby improving the accuracy of detection results of the trained neural network. In this way, the invention enables prediction of maintenance needs of industrial devices, avoids the occurrence of emergency repair events, and increases the uptime of the industrial devices, thereby improving the efficiency of industrial production.

Description

An abnormal vibration detection device for industrial equipment

technical field

The invention relates to the technical field of measurement, in particular to an abnormal vibration detection device of industrial equipment, which can be used for operation monitoring of industrial equipment.

Background technique

At present, intelligent automation of factory production has received more and more attention, because intelligent automation can not only improve production efficiency, but also reduce or eliminate serious losses caused by equipment downtime. Machine equipment imbalances, defects, loose fasteners, and other anomalies often translate into vibrations that reduce accuracy and raise safety concerns. In addition to performance and safety issues, if left unchecked, there will inevitably be a loss of productivity if equipment is shut down for repairs. Even small changes in equipment performance, which are often difficult to predict in time, can quickly translate into significant productivity losses.

In order to meet the inspection requirements of industrial equipment, at present most factories simply rely on manual inspection by technicians, but at this time, the labor consumption is huge and the reliability of inspection is difficult to guarantee. Some use simple piezoelectric sensors or hand-held data acquisition tools for device detection, but there are many limitations. On the one hand, human cooperation is still required, and on the other hand, these tools simply collect data and cannot detect whether an anomaly occurs or predict failure.

SUMMARY OF THE INVENTION

Embodiments of the present invention provide an abnormal vibration detection device for industrial equipment. In order to provide a basic understanding of some aspects of the disclosed embodiments, a brief summary is given below. This summary is not intended to be an extensive review, nor is it intended to identify key/critical elements or delineate the scope of protection of these embodiments. Its sole purpose is to present some concepts in a simplified form as a prelude to the detailed description that follows.

In a first aspect, an embodiment of the present invention provides a device for detecting abnormal vibration of industrial equipment. The device includes:

Industrial equipment, IoT node modules, and intelligent edge computing platforms; among them,

The industrial equipment is connected with the IoT node module, and the intelligent edge computing platform is communicated with the IoT node module, wherein,

The IoT node module detects and analyzes the vibration data of the industrial equipment to determine whether the vibration of the industrial equipment is abnormal.

Optionally, the IoT node module includes an embedded soft core, a neural network and a self-learning algorithm are built in the embedded soft core, and the self-learning algorithm is used to realize unsupervised online training of the neural network.

Optionally, the IoT node module includes a vibration sensor, a digital-to-analog converter, an acceleration unit, and a wireless communication unit; wherein, the vibration sensor, the digital-to-analog converter, the acceleration unit, and the wireless communication unit are electrically connected in sequence; wherein,

Vibration sensor and industrial equipment connection;

The wireless communication unit is communicatively connected to the intelligent edge computing platform.

Optionally, the acceleration unit includes an FPGA subunit;

The wireless communication unit includes a microcontroller with integrated wireless functionality; wherein,

The FPGA subunit includes an embedded soft core and a neural network; wherein,

The digital-to-analog converter is connected to the FPGA subunit through an integrated circuit bus; the FPGA subunit is connected to the microcontroller with integrated wireless function through a queue serial peripheral interface.

Optionally, the device further includes:

Alarm terminal; wherein, the alarm terminal and the intelligent edge computing platform are connected in communication.

Optionally, the vibration sensor is used to collect the vibration information of the industrial equipment to be detected in real time and send it to the digital-to-analog converter;

The digital-to-analog converter is used to convert the analog information of the vibration information into a digital signal and send it to the FPGA subunit in the acceleration unit through the integrated circuit bus;

The embedded soft core in the FPGA sub-unit is used to generate calculation results after calculation based on the pre-trained neural network, and send the calculation results to the microcontroller with integrated wireless function in the wireless communication unit through the queue serial peripheral interface;

The microcontroller with integrated wireless function is used for processing the calculation result through the mean square error method, generating a processing result, and determining whether the industrial equipment to be detected is abnormal based on the processing result.

Optionally, when the processing result is greater than or equal to a preset threshold, the microcontroller with integrated wireless function determines that the industrial equipment to be detected is abnormal; and when the processing result is less than the preset value, the microcontroller with integrated wireless function determines that the industrial equipment to be detected is abnormal. The device is not abnormal.

Optionally, when an abnormality occurs, the microcontroller with integrated wireless function generates early warning information, and sends the early warning information to the intelligent edge computing platform through the wireless communication unit;

The intelligent edge computing platform performs data processing and statistics on the early warning information, generates alarm information, and sends the alarm information to the alarm terminal.

Optionally, the embedded soft core generates a pre-trained neural network according to the following method, including:

Create model training components through self-learning algorithms;

Collect training data samples of multiple industrial equipment running normally, and input the training data samples into the neural network for unsupervised online training based on the model training component, and generate a pre-trained neural network after the training.

Optionally, after the pre-trained neural network is generated after the embedded soft core training is completed, the method further includes:

The embedded soft core downloads the weight file of the pre-trained neural network and writes the weight file into the embedded soft core.

The technical solutions provided by the embodiments of the present invention may include the following beneficial effects:

In the embodiment of the present invention, the device for detecting abnormal vibration of industrial equipment first collects the vibration information of the industrial equipment to be detected in real time through a vibration sensor and sends it to a digital-to-analog converter, and then converts the analog information of the vibration information into a digital signal through the digital-to-analog converter for transmission. to the FPGA sub-unit in the acceleration unit, and then based on the embedded soft core in the FPGA sub-unit to perform calculations based on the pre-trained neural network to generate calculation results and send the calculation results to the wireless communication through the queue serial peripheral interface The microcontroller with integrated wireless function in the unit is finally used to process the calculation result by the mean square error method through the microcontroller with integrated wireless function, generate the processing result, and determine whether the industrial equipment to be detected is abnormal based on the processing result. Because the present invention has built-in neural network and self-learning algorithm in the embedded soft core, it can realize online collection of normal operation data for unsupervised online training of the neural network, so that the detection result of the neural network after training is more accurate, so that the maintenance requirements of industrial equipment can be predicted. To avoid emergency maintenance events, maximize the uptime of industrial equipment, thereby improving industrial production efficiency.

It is to be understood that both the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the invention.

Description of drawings

The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the invention and together with the description serve to explain the principles of the invention.

1 is a schematic diagram of a device for detecting abnormal vibration of industrial equipment provided by an embodiment of the present invention;

2 is a schematic flowchart of a process for detecting abnormal vibration of industrial equipment according to an embodiment of the present invention;

3 is a schematic process diagram of an abnormal vibration detection process of an industrial equipment provided by an embodiment of the present invention;

Detailed ways

The following description and drawings sufficiently illustrate specific embodiments of the invention to enable those skilled in the art to practice them.

It should be understood that the described embodiments are only some, but not all, embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

Where the following description refers to the drawings, the same numerals in different drawings refer to the same or similar elements unless otherwise indicated. The implementations described in the illustrative examples below are not intended to represent all implementations consistent with the present invention. Rather, they are merely examples of apparatus and methods consistent with some aspects of the invention, as recited in the appended claims.

In the description of the present invention, it should be understood that the terms "first", "second" and the like are used for descriptive purposes only, and should not be construed as indicating or implying relative importance. For those of ordinary skill in the art, the specific meanings of the above terms in the present invention can be understood in specific situations. Furthermore, in the description of the present invention, unless otherwise specified, "a plurality" means two or more. "And/or", which describes the association relationship of the associated objects, means that there can be three kinds of relationships, for example, A and/or B, which can mean that A exists alone, A and B exist at the same time, and B exists alone. The character "/" generally indicates that the associated objects are an "or" relationship.

So far, for the detection of abnormal vibration of industrial equipment, most factories simply rely on manual inspection by technicians, but at this time, the labor consumption is huge and the reliability of detection is difficult to guarantee. Some use simple piezoelectric sensors or hand-held data acquisition tools for device detection, but there are many limitations. On the one hand, human cooperation is still required, and on the other hand, these tools simply collect data and cannot detect whether an anomaly occurs or predict failure. Therefore, the present invention provides a device, method, storage medium and electronic device for detecting abnormal vibration of industrial equipment, so as to solve the problems existing in the above-mentioned related technical problems. In the technical solution provided by the present invention, since the present invention has built-in neural network and self-learning algorithm in the embedded soft core, it can realize online collection of normal operation data for unsupervised online training of the neural network, so that the neural network detection result after training is more accurate , so that the maintenance requirements of industrial equipment can be predicted to avoid emergency maintenance events, so as to maximize the normal operation time of industrial equipment, thereby improving industrial production efficiency. The following uses exemplary embodiments to describe in detail.

Please refer to FIG. 1. FIG. 1 is a schematic diagram of a device for detecting abnormal vibration of industrial equipment according to an embodiment of the present invention. The device includes industrial equipment, an IoT node module, and an intelligent edge computing platform; wherein, the industrial equipment and the IoT node The module is connected, and the intelligent edge computing platform is connected to the IoT node module in communication, wherein the IoT node module detects and analyzes the vibration data of the industrial equipment to determine whether the vibration of the industrial equipment is abnormal.

Further, the IoT node module includes an embedded soft core, a neural network and a self-learning algorithm are built in the embedded soft core, and the self-learning algorithm is used to realize unsupervised online training of the neural network.

It should be noted that the IoT node module includes an embedded soft core, a neural network and a self-learning algorithm are built in the embedded soft core, and the self-learning algorithm is used to realize unsupervised online training of the neural network.

Specifically, industrial equipment such as: industrial fans, mixers, turbine generators, motors, water pumps, and gearboxes. The intelligent edge computing platform is used to receive data from IoT nodes for data processing and statistics. The alarm terminal includes mobile phone alarm and big data platform alarm, and real-time alarm when the vibration is abnormal.

Further, the IoT node module includes a vibration sensor, a digital-to-analog converter, an acceleration unit, and a wireless communication unit; wherein, the vibration sensor, the digital-to-analog converter, the acceleration unit, and the wireless communication unit are electrically connected in sequence; wherein, the vibration sensor and the industrial equipment connection; the wireless communication unit communicates with the intelligent edge computing platform.

Specifically, vibration sensors can detect a variety of industrial equipment, and are used to collect operational data information generated by the industrial equipment. Among them, digital-to-analog converter, also known as D/A converter, referred to as DAC, is a device that converts digital quantities into analog.

Further, the acceleration unit includes an FPGA subunit; wherein, the wireless communication unit includes a microcontroller integrating wireless functions; wherein, the FPGA subunit includes an embedded soft core and a neural network; The FPGA subunit is connected; the FPGA subunit is connected with the microcontroller with integrated wireless function through a queue serial peripheral interface.

Further, the device further includes: an alarm terminal; wherein, the alarm terminal is connected in communication with the intelligent edge computing platform.

In a possible implementation, the six-axis acceleration sensor first collects the digital signal corresponding to the vibration information of the industrial equipment and sends it to the embedded soft core (the microprocessor IP of the FPGA) through the IIC (Inter-Integrated Circuit, integrated circuit bus). Core), in which the microprocessor IP core of FPGA runs fast, occupies less resources, and has strong configurability. In the microprocessor IP core of the FPGA, there is a model of the industrial equipment to be tested created by the Long Short-Term Memory Neural Network (LSTM, Long Short-Term Memory) algorithm. The neural network is implemented using 16-bit fixed-point numbers, and the trained model is compressed by sparse matrix. The algorithm performs parallel computing to realize the deployment of a two-layer 64-point LSTM network, which meets the requirements of real-time computing monitoring. Finally, the computing results are sent to the microcontroller with integrated wireless function through the queue serial peripheral interface.

Further, in the microcontroller with integrated wireless function, the mean square error is used in this module to judge whether it is abnormal, if the alarm information is higher than the threshold, the alarm information will be transmitted to the intelligent edge computing server through the wireless network, and if it is lower than the threshold, it will be silent.

Referring to FIG. 2 , an embodiment of the present invention provides a schematic flowchart of an industrial equipment abnormal vibration detection process applied to an industrial equipment abnormal vibration detection apparatus. As shown in FIG. 2 , the detection process of the embodiment of the present invention may include the following steps:

S101, the vibration sensor is used to collect the vibration information of the industrial equipment to be detected in real time and send it to the digital-to-analog converter;

Among them, the vibration sensor is a sensor component that detects the vibration of the equipment. Vibration information is the data information collected when industrial equipment vibrates during operation. It should be noted that the vibration sensor is preferably a six-axis acceleration sensor. The vibration sensor (six-axis acceleration sensor) transmits the data to the digital-to-analog conversion unit, and then transmits it to the FPGA through the integrated line bus, and the FPGA transmits it to the wireless communication unit through the queue serial peripheral interface (the integrated wireless function is the controller).

In a possible implementation manner, when the abnormal vibration detection device of industrial equipment detects the abnormality of industrial equipment, the user first connects the industrial equipment to be detected with the abnormal vibration detection device of industrial equipment, and starts the abnormal vibration detection of industrial equipment after the connection is successful. The device performs abnormality detection on the industrial equipment to be detected, and collects the vibration information of the industrial equipment to be detected in real time through the vibration sensor during detection.

S102, the digital-to-analog converter is used to convert the analog information of the vibration information into a digital signal and send it to the FPGA subunit in the acceleration unit;

Among them, digital-to-analog converter, also known as D/A converter, referred to as DAC, is a device that converts digital quantities into analog.

In a possible implementation manner, after the vibration source of the industrial equipment generates vibration, the vibration data is collected by the vibration sensor and input to the digital-to-analog converter, and the digital-to-analog converter converts the analog signal into a digital signal.

S103, the embedded soft core in the FPGA sub-unit is used to perform computation based on the pre-trained neural network to generate a computation result and send the computation result to the micro-controller with integrated wireless function in the wireless communication unit through the queue serial peripheral interface device;

Usually, the pre-trained model of the industrial equipment to be inspected in the embedded soft core is generated after collecting normal operation data online for unsupervised online training.

Further, during the training of the industrial equipment model to be tested, the long-short-term memory neural network is first used to create the industrial equipment model to be tested, and then the model training component is created by the self-learning algorithm, and finally a plurality of training data samples of the normal operation of the industrial equipment are collected, and based on the The model training component inputs the training data samples into the model of the industrial equipment to be tested for training and performs unsupervised online training, and generates a pre-trained model of the industrial equipment to be tested after the training.

Further, after the model training is completed, the weight file of the pre-trained industrial equipment model to be detected is downloaded, and the weight file is written into the embedded soft core. At this time, the embedded soft core has the detection capability.

In a possible implementation manner, when the embedded soft core has the detection capability, the digital signal of the industrial equipment to be detected is sent to the embedded soft core through the integrated circuit bus, and based on the pre-trained to-be-detected soft core Detect the weight file in the industrial equipment model and perform the calculation to generate the calculation result.

It should be noted that the vibration sensor, digital-to-analog converter, acceleration unit and wireless communication are all completed in the IoT node. Compared with similar products on the market, neural network and self-learning algorithm are added, which does not require manual calibration of training data, and realizes self-learning with the cooperation of edge computing and IOT nodes. At the same time passive wireless, easy to deploy.

S105, the microcontroller with integrated wireless function is used to process the calculation result through the mean square error method, generate the processing result, and determine whether the industrial equipment to be detected is abnormal based on the processing result.

Among them, mean-square error (MSE) is a measure that reflects the degree of difference between the estimator and the estimator.

In a possible implementation manner, after the calculation result is obtained based on step S104, the calculation result is used as the parameter of the mean square error algorithm, and the processing result is obtained after the second calculation. According to the processing result, it can be determined whether the industrial equipment to be detected appears abnormal.

For example, when the processing result is greater than or equal to the preset threshold, it is determined that the industrial equipment to be detected is abnormal; and when the processing result is less than the preset value, it is determined that the industrial equipment to be detected is not abnormal.

Further, when an abnormality occurs, early warning information is generated, and the early warning information is sent to the intelligent edge computing platform through the microcontroller with integrated wireless function. The alarm information is sent to the alarm terminal.

For example, as shown in FIG. 3, FIG. 3 is an architecture diagram of an abnormal vibration detection of industrial equipment provided by an embodiment of the present invention. The architecture includes a vibration source (ie, the industrial equipment to be detected), an IoT node (including a vibration sensor, a digital-to-analog converter, and a digital-to-analog converter). (A/D unit), acceleration unit and wireless communication), intelligent edge computing platform and alarm platform (alarm platform includes abnormal vibration mobile phone APP alarm and abnormal vibration big data platform alarm).

Further, the acceleration unit includes an FPGA subunit, and the FPGA subunit includes an embedded soft core and an industrial equipment model to be detected.

An example diagram of a scene in an embodiment of the present invention is shown in FIG. 4 .

The present invention further provides a computer-readable medium on which program instructions are stored, and when the program instructions are executed by a processor, implement the methods for detecting abnormal vibration of industrial equipment provided by the above method embodiments.

The present invention also provides a computer program product containing instructions, which, when running on a computer, enables the computer to execute the method for detecting abnormal vibration of industrial equipment in each of the above method embodiments.

Those of ordinary skill in the art can understand that the realization of all or part of the processes in the methods of the above embodiments can be accomplished by instructing the relevant hardware through a computer program, and the program can be stored in a computer-readable storage medium, and when the program is executed , which may include the processes of the above-mentioned method embodiments. Wherein, the storage medium can be a magnetic disk, an optical disk, a read-only storage memory or a random storage memory, and the like.

The above disclosures are only preferred embodiments of the present invention, and of course, the scope of the rights of the present invention cannot be limited by this. Therefore, equivalent changes made according to the claims of the present invention are still within the scope of the present invention.

Claims

A device for detecting abnormal vibration of industrial equipment, characterized in that the device comprises:

Industrial equipment, IoT node modules, and intelligent edge computing platforms; among them,

The industrial equipment is connected to the IoT node module, and the intelligent edge computing platform is communicatively connected to the IoT node module, wherein,

The IoT node module detects and analyzes the vibration data of the industrial equipment to determine whether the vibration of the industrial equipment is abnormal.
A device for detecting abnormal vibration of industrial equipment according to claim 1, characterized in that

The IoT node module includes an embedded soft core, and a neural network and a self-learning algorithm are built in the embedded soft core, and the self-learning algorithm is used to realize the unsupervised online training of the neural network.
A device for detecting abnormal vibration of industrial equipment according to claim 1 or 2, characterized in that:

The IoT node module includes a vibration sensor, a digital-to-analog converter, an acceleration unit, and a wireless communication unit; wherein, the vibration sensor, the digital-to-analog converter, the acceleration unit, and the wireless communication unit are electrically connected in sequence; wherein,

the vibration sensor is connected to the industrial equipment;

The wireless communication unit is in communication connection with the intelligent edge computing platform.
A device for detecting abnormal vibration of industrial equipment according to claim 3, characterized in that:

The acceleration unit includes an FPGA subunit;

The wireless communication unit includes a microcontroller with integrated wireless functionality; wherein,

The FPGA subunit includes an embedded soft core and a neural network; wherein,

The digital-to-analog converter is connected to the FPGA subunit through an integrated circuit bus; the FPGA subunit is connected to the microcontroller with integrated wireless function through a queue serial peripheral interface.
The device for detecting abnormal vibration of industrial equipment according to claim 1, wherein the device further comprises:

An alarm terminal; wherein, the alarm terminal is connected in communication with the intelligent edge computing platform.
A device for detecting abnormal vibration of industrial equipment according to claim 4, characterized in that:

The vibration sensor is used to collect the vibration information of the industrial equipment to be detected in real time and send it to the digital-to-analog converter;

The digital-to-analog converter is used to convert the analog information of the vibration information into a digital signal and send it to the FPGA subunit in the acceleration unit through the integrated circuit bus;

The embedded soft core in the FPGA sub-unit is used to generate calculation results after calculation based on the pre-trained neural network, and send the calculation results to the microcontroller with integrated wireless function in the wireless communication unit through the queue serial peripheral interface;

The microcontroller with integrated wireless function is used for processing the calculation result through the mean square error method, generating a processing result, and determining whether the industrial equipment to be detected is abnormal based on the processing result.
A device for detecting abnormal vibration of industrial equipment according to claim 6, characterized in that:

When the processing result is greater than or equal to a preset threshold, the microcontroller with integrated wireless function determines that the industrial equipment to be detected is abnormal; and when the processing result is less than the preset value, the microcontroller with integrated wireless function determines that the There is no abnormality in the industrial equipment to be tested.
A device for detecting abnormal vibration of industrial equipment according to claim 7, characterized in that:

When an abnormality occurs, the microcontroller with integrated wireless function generates early warning information, and sends the early warning information to the intelligent edge computing platform through the wireless communication unit;

The intelligent edge computing platform performs data processing and statistics on the early warning information, generates alarm information, and sends the alarm information to the alarm terminal.
The device for detecting abnormal vibration of industrial equipment according to claim 6, wherein the embedded soft core generates a pre-trained neural network according to the following method, comprising:

Create model training components through self-learning algorithms;

Collecting a plurality of training data samples of normal operation of industrial equipment, and inputting the training data samples into the neural network based on the model training component for unsupervised online training, and generating a pre-trained neural network after the training.
The device for detecting abnormal vibration of industrial equipment according to claim 9, characterized in that, after the embedded soft core training is completed and the pre-trained neural network is generated, it further comprises:

The embedded soft core downloads the weight file of the pre-trained neural network and writes the weight file into the embedded soft core.