WO2012088707A1 - 用于设备故障检测的智能检测系统及检测方法 - Google Patents

用于设备故障检测的智能检测系统及检测方法 Download PDF

Info

Publication number
WO2012088707A1
WO2012088707A1 PCT/CN2010/080579 CN2010080579W WO2012088707A1 WO 2012088707 A1 WO2012088707 A1 WO 2012088707A1 CN 2010080579 W CN2010080579 W CN 2010080579W WO 2012088707 A1 WO2012088707 A1 WO 2012088707A1
Authority
WO
WIPO (PCT)
Prior art keywords
data
detection system
intelligent detection
board
central processing
Prior art date
Application number
PCT/CN2010/080579
Other languages
English (en)
French (fr)
Inventor
谭民
赵晓光
梁自泽
李恩
Original Assignee
中国科学院自动化研究所
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 中国科学院自动化研究所 filed Critical 中国科学院自动化研究所
Priority to US13/976,882 priority Critical patent/US20140298099A1/en
Priority to PCT/CN2010/080579 priority patent/WO2012088707A1/zh
Priority to CN2010800706166A priority patent/CN103250107A/zh
Publication of WO2012088707A1 publication Critical patent/WO2012088707A1/zh

Links

Classifications

    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F11/00Error detection; Error correction; Monitoring
    • G06F11/07Responding to the occurrence of a fault, e.g. fault tolerance
    • G06F11/0703Error or fault processing not based on redundancy, i.e. by taking additional measures to deal with the error or fault not making use of redundancy in operation, in hardware, or in data representation
    • G06F11/079Root cause analysis, i.e. error or fault diagnosis
    • GPHYSICS
    • G05CONTROLLING; REGULATING
    • G05BCONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
    • G05B19/00Programme-control systems
    • G05B19/02Programme-control systems electric
    • G05B19/18Numerical control [NC], i.e. automatically operating machines, in particular machine tools, e.g. in a manufacturing environment, so as to execute positioning, movement or co-ordinated operations by means of programme data in numerical form
    • G05B19/406Numerical control [NC], i.e. automatically operating machines, in particular machine tools, e.g. in a manufacturing environment, so as to execute positioning, movement or co-ordinated operations by means of programme data in numerical form characterised by monitoring or safety
    • GPHYSICS
    • G05CONTROLLING; REGULATING
    • G05BCONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
    • G05B2219/00Program-control systems
    • G05B2219/30Nc systems
    • G05B2219/37Measurements
    • G05B2219/37226Monitor condition of spindle, tool holder, transmit to nc controller
    • GPHYSICS
    • G05CONTROLLING; REGULATING
    • G05BCONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
    • G05B2219/00Program-control systems
    • G05B2219/30Nc systems
    • G05B2219/37Measurements
    • G05B2219/37337Noise, acoustic emission, sound
    • GPHYSICS
    • G05CONTROLLING; REGULATING
    • G05BCONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
    • G05B2219/00Program-control systems
    • G05B2219/30Nc systems
    • G05B2219/37Measurements
    • G05B2219/37431Temperature
    • GPHYSICS
    • G05CONTROLLING; REGULATING
    • G05BCONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
    • G05B2219/00Program-control systems
    • G05B2219/30Nc systems
    • G05B2219/37Measurements
    • G05B2219/37434Measuring vibration of machine or workpiece or tool

Definitions

  • the invention relates to an intelligent detection system for detecting equipment operation failure, and adopts intelligent sensing technology and embedded computer science technology. Specifically, it relates to an intelligent detection system and a detection method for device fault detection. Background technique
  • the object of the present invention is to provide an intelligent detection system and a detection method for equipment failure detection, which realize real-time mechanical failure of large equipment based on embedded technology. , intelligent detection and diagnosis.
  • an intelligent detection system for device fault detection is provided, and the smart detection system is externally connected with data for collecting an operation state of the device to be detected.
  • a plurality of sensors the smart detection system comprising: a central processing board 2 having a central processing unit and a plurality of data interfaces coupled to the central processing unit; a data acquisition board 3 coupled to one or more of the plurality of sensors For processing data collected by the sensor; a synchronous communication board 4 for synchronizing communication between the central processing board 2 and the data acquisition board 3;
  • the plurality of connection plug-ins 1 are connected to the central processing board 2, the data acquisition board 3, and the synchronous communication board 4 to implement data transmission.
  • the central processing unit is configured to analyze data collected by the plurality of sensors. When the collected data exceeds the set limit, it indicates that the status of the device to be detected is abnormal, and an alarm message is issued. When the collected data does not exceed the set limit, it indicates that the status of the device to be tested is normal.
  • the plurality of sensors comprise a vibration sensor, an acoustic sensor and a temperature sensor.
  • the central processing unit analyzes temperature data collected by the temperature sensor, and sends an alarm message when the temperature data exceeds a preset upper and lower limit.
  • the central processing unit analyzes the vibration signal collected by the vibration sensor to obtain the vibration amplitude and the vibration frequency characteristic; and presets the upper and lower limits of the vibration amplitude, when the acquired vibration amplitude exceeds the set Alarm is given when the upper and lower limits are fixed; the frequency characteristic of the vibration signal under normal conditions is preset, and the collected vibration frequency characteristic is compared with the frequency characteristic, and an alarm message is issued when an abnormality occurs.
  • the central processing unit analyzes the acoustic signal collected by the acoustic sensor to obtain a sound pressure level value and a sound frequency characteristic; and presets a maximum sound pressure level value, when the collected sound pressure level value exceeds The maximum sound pressure level value is used to issue an alarm message; the frequency characteristic of the acoustic signal under normal conditions is preset, and the collected acoustic frequency characteristic is compared with the frequency characteristic, and an alarm message is issued when an abnormality occurs.
  • the plurality of connection plug-ins 1 adopt a PC104 bus structure.
  • the multiple data interfaces include a serial port, a 485 bus interface, a CAN bus interface, a network interface, and an optoelectronic conversion interface.
  • the temperature sensor is connected to the 485 bus interface, and the vibration sensor and the acoustic sensor are connected to the data acquisition board 3.
  • the data acquisition board 3 is a multi-channel high-speed data acquisition board.
  • the smart detection system further includes a self-operating state monitoring module, configured to monitor whether the smart detection system has a fault, and output an alarm message when a fault occurs.
  • a self-operating state monitoring module configured to monitor whether the smart detection system has a fault, and output an alarm message when a fault occurs.
  • the working status of the monitored intelligent detection system itself includes: the plurality of sensors are not collected, the data collection board 3 is abnormal, and the multiple data interface connections are interrupted.
  • an intelligent detection method for an intelligent detection system is provided, the smart detection system being connected with a plurality of sensors for collecting operational status data of a device to be detected, the intelligent detection system including central processing a board 2, a data acquisition board 3, a synchronous communication board 4, and a plurality of connection plug-ins 1, through which the central processing board 2, the data acquisition board 3, and the synchronous communication board 4 are connected to realize data transmission.
  • the intelligent detection method comprises the following steps: a system initialization step of initializing each device and interface of the intelligent detection system to make it enter a ready state; a working mode setting step, setting a working mode of the intelligent detection system as a main control Or slave controller; working timing setting step, setting working timing and task priority of each device and interface of the intelligent detecting system; data collecting step, the intelligent detecting system according to the set working sequence and task The priority starts to perform a data collection task; the data analysis step, the wisdom The detection system analyzes the collected data, and determines whether the device to be detected is in a normal state according to the data analysis result.
  • the working timing setting step includes setting a working sequence and a task priority of data collection, data processing, and data transmission of the plurality of sensors.
  • the data transmission task of the network interface has the highest priority
  • the data acquisition task of the temperature sensor has the highest priority
  • the data acquisition task of the vibration sensor and the acoustic sensor has the highest priority
  • the data analysis step further includes the following steps: When the data exceeds the set limit, it indicates that the status of the device to be detected is abnormal, and an alarm message is issued; when the collected data does not exceed the set limit, it indicates that the status of the device to be detected is normal, returning to the above-mentioned working mode setting step, and continuing A round of testing.
  • the upper and lower limits of the temperature are preset, and when the temperature data collected by the temperature sensor exceeds the upper and lower limits set, an alarm message is issued.
  • the data analysis step analyzes the vibration signal obtained by the vibration sensor to obtain the vibration amplitude and the vibration frequency characteristic; and presets the upper and lower limits of the vibration amplitude, when the acquired vibration amplitude exceeds the set upper and lower limits. Perform an alarm; preset the frequency characteristic of the vibration signal under normal conditions, and compare the acquired vibration frequency characteristic with the frequency characteristic In comparison, an alarm message is issued when an abnormality occurs.
  • the data analysis step analyzes the acoustic signal collected by the acoustic sensor to obtain the sound pressure level value and the sound frequency characteristic; and presets the maximum sound pressure level value, when the collected sound pressure level value exceeds the maximum sound Alarm is given when the pressure level is used; the frequency characteristic of the acoustic signal under normal conditions is preset, and the collected acoustic frequency characteristic is compared with the frequency characteristic, and an alarm message is issued when an abnormality occurs.
  • the intelligent detection system further includes a self-operating state monitoring step, configured to monitor whether the intelligent detection system itself has a fault, and output an alarm message when a fault occurs.
  • the working state of the monitored intelligent detection system itself includes: the plurality of sensors do not collect data, the data acquisition board 3 works abnormally, and multiple data interface connections connected to the central processing board 2 are interrupted. .
  • the temperature sensor, the vibration sensor, and the acoustic sensor are installed at the detecting portion of the device, and the temperature of the rotating member, the vibration of the fastening member, and the acoustics during the processing are detected in real time during the operation of the device.
  • the signal changes, so that the hidden troubles or fault points can be found in time, and the alarm is repaired in time to avoid equipment damage. Therefore, the efficiency and accuracy of equipment fault detection are greatly improved, and an effective guarantee for the safe operation of the equipment is provided.
  • FIG. 1 is a schematic structural view of an intelligent detection system according to an embodiment of the present invention
  • FIG. 2 is a schematic structural view of a central processing board shown in FIG.
  • Figure 3 shows the outer casing of the intelligent detection system of the present invention
  • FIG. 4 is a flow chart showing an intelligent detecting method of the intelligent detecting system according to the present invention
  • FIG. 5 is a topological structural view showing a BP neural network according to an embodiment of the present invention
  • Figure 6 shows the vibration frequency characteristics of the device to be tested under normal conditions
  • Figure 7 shows the vibration frequency characteristics of the device to be tested in the fault state.
  • FIG. 1 is a schematic structural diagram of an intelligent detection system according to an embodiment of the present invention.
  • the intelligent detecting system includes a central processing board 2, a data collecting board 3, and a synchronous communication board 4.
  • the data bus of the central processing board 2 adopts a structure compatible with the PC104 bus.
  • the data bus is usually implemented by a row of connectors, one end of which is a jack, and the other end is a connector 1 with a pin.
  • the central processing board 2, the data acquisition board 3 and the synchronous communication board 4 of the present invention each have such a connection plug 1 so that the boards can be connected together to facilitate data transfer.
  • the central processing board 2 preferably employs an embedded central processing board on which is mounted a central processing unit for controlling the operation of the entire intelligent detection system.
  • the synchronous communication board 4 is connected between the central processing board 2 and the data acquisition board 3, and is provided with a timing adjustment circuit, and the synchronous communication board 4 communicates with the data processing board 3 through the communication between the central processing board 2 and the data acquisition board 3. Timing adjustments are made to keep the communication between the two in sync.
  • the synchronizing communication board 4 is provided with a connection plug 1 and is electrically connected to the central processing board 2 and the data acquisition board 3 via the connection plug 1.
  • the data acquisition board 3 includes a plurality of data acquisition channels and registers, the registers are used to store the data of the current acquisition channel, and the central processing unit reads the data in the registers through the PC104 bus.
  • the data acquisition board 3 of the present invention preferably uses a multi-channel high-speed data acquisition board, such as DM6430HR-l o
  • Figure 2 is a schematic view showing the structure of the central processing board shown in Figure 1.
  • the central processing board mainly includes a central processing unit, which preferably employs an ARM9 type microprocessor chip.
  • the central processing unit of the present invention can expand the memory for storing program run files and temporary data.
  • the memory is preferably FLASH memory (i.e., flash memory) to ensure a high data read rate.
  • the memory is 32M capacity.
  • the intelligent detection system of the invention adopts the PC104 bus structure, and the central processing unit can be connected to external devices such as the data acquisition board 3, the synchronous communication board 4 through the PC104 bus, and control the operation of these external devices.
  • the central processing board also integrates multiple data interfaces, which are connected between the external device and the central processing unit, and serve as a communication interface for data transmission and command.
  • These data interfaces correspond to different data interface standards, such as serial port, 485 bus interface, CAN bus interface (CAN bus is a bus standard), network interface, photoelectric conversion interface, etc.
  • the serial port is preferably a standard 232 serial port.
  • state data reflecting the real-time operating state of the device to be detected such as a vibration sensor, an acoustic sensor, a temperature sensor, a pressure sensor, an acceleration sensor, and the like, are collected by a plurality of sensors.
  • Different sensor output data types are different, so it is necessary to transfer these data to the central processing unit on the central processing board 2 using different data transmission interfaces.
  • different sensors are connected to different standard interfaces, S ⁇ 232 interface, 485 interface, CAN interface, network interface, and photoelectric conversion interface.
  • the temperature sensor is connected to the 485 interface, and the vibration sensor and the acoustic sensor are directly connected to the data acquisition board 3 for collecting state parameters reflecting the real-time operating state of the device to be detected, such as vibration, noise, etc. .
  • state parameters reflecting the real-time operating state of the device to be detected such as vibration, noise, etc.
  • 16 temperature sensors are connected using the 485 interface, and one vibration sensor and four acoustic sensors are connected using the high speed data acquisition board HR6430-1.
  • the intelligent detection system of the present invention is equipped with a power supply and corresponding power supply circuitry for providing a constant voltage, such as a constant voltage of 3V, 5V, 12V, for the central processing unit, the data acquisition board, and externally connected sensors.
  • a constant voltage such as a constant voltage of 3V, 5V, 12V
  • Figure 3 shows the housing of the intelligent detection system of the present invention.
  • the intelligent detecting system of the present invention is packaged in a casing 3-1 for accommodating and protecting components of an intelligent detecting system disposed therein.
  • the two sides of the outer casing 3-1 are respectively provided with openings and sockets for power supply, sensor leads and external communication interfaces.
  • the front of the casing 3-1 is provided with two openings, two D-type sockets (a universal socket, the shape of the letter D), one power socket 3-2, 220v power supply, and the other sensor Lead plug 3-3.
  • One side of the outer casing 3-1 is provided with four openings, which are respectively connected to the RS232 interface 3-4, the 485 interface 3-5, the CAN bus interface 3-6 and the network interface 3-7, and these interfaces are used for connecting external lines to realize data. Transmission of information.
  • the smart detection method of the present invention includes the following steps:
  • System initialization steps That is, the various devices and interfaces of the intelligent detection system are initialized, so that the system Enter the ready state. Specifically, the central processing board 2, the data acquisition board 3, and the synchronous communication board 4 are detected, and it is determined whether the devices and devices connected through the external interface are in a normal state.
  • the working mode of the intelligent detection system is set as the master controller or the slave controller.
  • the operation mode of the intelligent detection system is set as the master controller.
  • various control commands are generated by the intelligent detection system to direct other external devices to work.
  • the operation mode of the smart detection system is set as the slave controller.
  • the intelligent detection system acts as an external device of the external main controller, and responds to various control commands sent by the external main controller.
  • the workflow is given by taking the smart detection system as the main controller as an example.
  • the setting of the working mode can be automatically performed according to the preset setting, or can be manually set by the operator during the operation of the intelligent detecting system.
  • Work timing setting steps According to the type of device to be tested, set the working sequence and task priority of each device and interface of the intelligent detection system to keep it in a normal working state. For example, if the device to be tested is sensitive to temperature parameters, the data acquisition priority of the set temperature sensor is high; if it is sensitive to vibration or noise, the data acquisition priority of the vibration sensor or acoustic sensor is set high.
  • the operation timing setting step includes, for example, setting data collection, data processing, and operation timing and task priority of the data transmission of the plurality of sensors.
  • the data transmission task of the network interface has the highest priority
  • the temperature sensor connected by the 485 bus has the highest priority
  • the vibration sensor and the acoustics connected through the data acquisition board are the highest.
  • the sensor's data collection task has the lowest priority.
  • the intelligent detection system starts the data collection task according to the set working timing and task priority. That is, the data reflecting the operating state of the device to be detected is collected by a sensor disposed at each detecting position of the device to be detected.
  • the intelligent detection system analyzes the collected data, and determines whether the device to be detected is in a normal state according to the result of the data analysis. Specifically, when the collected data exceeds the set limit, the state of the device to be detected is abnormal, and an alarm message is issued; when the collected data does not exceed the set limit, the state of the device to be detected is normal, and the working mode is returned to the previous working mode. Steps to continue the next round of testing.
  • a temperature sensor, a vibration sensor, and an acoustic sensor and an analysis method for collecting data for each sensor is as follows.
  • Temperature data Set the upper and lower limits of the temperature. When the temperature value collected by the temperature sensor exceeds the upper and lower limits set, an alarm message is issued.
  • Vibration data The vibration sensor collects the vibration signal of the target device, and when the vibration signal is abnormal, an alarm message is issued.
  • the analysis of the vibration signal includes the amplitude of the vibration and/or the frequency of the vibration.
  • the upper and lower limits of the vibration amplitude can be set in advance.
  • the amplitude is processed by the time domain to obtain the amplitude, and when the amplitude exceeds the upper and lower limits of the set amplitude, an alarm is issued.
  • the frequency characteristic of the vibration signal under normal conditions can be set in advance.
  • the Fourier transform algorithm or the wavelet analysis algorithm is used to convert the vibration signal into a periodic frequency domain value, and compare it with the frequency characteristic of the vibration signal under the normal condition set in advance, and issue an alarm message when an abnormality occurs. .
  • Acoustic data The processing of the acoustic signal is similar to the vibration signal.
  • the maximum sound pressure level value and the frequency characteristic of the acoustic signal under normal conditions are set.
  • the sound pressure level value is obtained by time domain processing of the acoustic signal. When the sound pressure level value exceeds the set value The alarm is given when the maximum sound pressure level value is set.
  • the frequency characteristic of the acoustic signal is obtained by the Fourier transform algorithm or the wavelet analysis algorithm, and compared with the frequency characteristic of the acoustic signal in a predetermined normal condition, and an alarm message is issued when an abnormality occurs.
  • the sensor can be set as follows. Temperature sensor: Two temperature sensors are mounted on the sleeve of the spindle. The upper limit of the temperature sensor is set to 80 degrees Celsius and the lower limit is set to 10 degrees Celsius.
  • the device status is normal; otherwise, if the upper and lower limits of the value range are exceeded, an alarm message is issued.
  • Vibration sensor Two vibration sensors are installed in the axial and radial directions of the main shaft to detect the axial and radial vibration of the main shaft.
  • the upper and lower limits of the radial and axial vibration amplitudes of the main shaft are set to ⁇ 0.5mm. .
  • the table Indicates that the device status is normal; otherwise, if the upper and lower limits of the value range are exceeded, an alarm message is issued.
  • Acoustic sensor Four acoustic sensors are installed around the main shaft. The probe is facing the main shaft. It is used to collect the sound from the main shaft. The upper limit of the sound pressure level of the main shaft is set to 94DB, and the lower limit is set to 50DB.
  • the sound pressure level value collected by the acoustic sensor is between 50DB and 94DB, it means the device status is normal; otherwise, if the upper and lower limits of the value range are exceeded, an alarm message is issued.
  • the vibration signal is taken as an example to describe in detail the acquisition and processing of the frequency characteristics of the vibration signal.
  • the vibration sensor collects the vibration signal (in the form of a vibration waveform) during normal operation of the spindle.
  • the sampling frequency is 10 kHz and the sampling output is ⁇ 5 volts.
  • the 4th-order wavelet of the Daubenchies wavelet series is used to perform 3-layer wavelet packet decomposition, and the frequency band of 8 components is obtained.
  • the total energy is obtained for the signals of the eight frequency bands, the feature vector is constructed with the energy as the element, and the feature vector is normalized to obtain the normalized feature vector.
  • the normalized eigenvectors of the vibration signals obtained by a single sampling are shown in Table 1 below:
  • the above process is repeated 14 times, and the vibration signals of 14 sets of spindles during normal operation are obtained.
  • 14 sets of normalized feature vectors are obtained, as shown in Table 2 above.
  • Fig. 5 is a diagram showing the topology of a BP neural network in accordance with an embodiment of the present invention.
  • a BP (back propagation) neural network (error back propagation neural network) is preliminarily established in the intelligent detection system of the present invention, and the neural network includes an input layer, an intermediate layer, and an output layer.
  • the input layer is used to input a normalized feature vector obtained by real-time acquisition of the signal through wavelet transform, including 8 neuron nodes, and each neuron node corresponds to one of 8 components of each set of feature vectors; It includes 4 neuron nodes for processing the input data of the input layer to improve the calculation accuracy of the neural network; the output layer includes 2 neuron nodes, and the output value of the output layer is 0 or 1, respectively indicating that the state is normal or The status is abnormal (that is, there is a fault).
  • the middle layer neuron selects tansig (S-type tangent function) as the activation function
  • the output layer neurons select logsig (S-type logarithmic function) as the activation function.
  • tansig S-type tangent function
  • logsig S-type logarithmic function
  • the design requirements of the neural network of the present invention are: The neural network training is up to 20000 iterations, and the output error is less than 0.002.
  • the vibration frequency characteristics of the machine spindle under normal conditions can be obtained, as shown in Fig. 6.
  • Figure 6 shows the vibration frequency characteristics of the device to be tested under normal conditions.
  • the waveforms of the eight components of each group of vibration signals in the normal state are displayed.
  • the classification number of the output fault classification is 0, indicating that the state is normal.
  • the vibration sensor collects the vibration signal of the machine tool spindle in real time, and processes the signal according to the above steps.
  • the abnormal vibration signal passes the wavelet change
  • the output of the neural network is 1, indicating that there is a fault, and the system issues an alarm message.
  • Figure 7 shows the vibration frequency characteristics of the device to be tested in the fault state.
  • the waveforms of the eight components of each group of vibration signals in the fault state are shown.
  • the waveforms of the eight components of each group of vibration signals in the fault state are significantly different from the waveforms of the corresponding components in the normal state.
  • the classification number of the output fault classification is 1, indicating that there is a fault, and the system will issue an alarm message.
  • the number of repetitions 14 in this embodiment is not fixed, but may be more or less depending on the actual situation.
  • the design requirements can be met by repeating 14 times: The neural network training is up to 20000 times and the output error is less than 0.002.
  • the processing of the acoustic signal is the same as that described above for the vibration signal.
  • the acoustic signal is also subjected to wavelet transform to obtain a normalized eigenvector.
  • the eigenvectors in the normal state are used to train the neural network to obtain the frequency characteristics of the acoustic signal under normal conditions.
  • the network output is 1, indicating that a fault has been detected and the system issues an alarm message.
  • the workflow is similar to the setting procedure of the above-described master controller, except that the work timing setting is performed by the external master controller.
  • the intelligent detection system of the present invention is further provided with a self-operating state monitoring module (not shown) for monitoring whether the working state of the intelligent detecting system itself is normal.
  • the workflow of the self-operating state monitoring module operates independently of the above-described process of detecting the working state of the device to be tested.
  • the fault state or abnormal state of the intelligent detection system of the present invention includes, for example, the following conditions:
  • the sensor connected to the intelligent detection system through each data interface (such as RS232 interface, 485 bus interface) has no data, and is connected to the central processing unit through the PC104 bus.
  • the data acquisition board works abnormally, and the network communication connection is interrupted.
  • the intelligent detection system issues an alarm message in time. In this way, reliable operation and timely maintenance of the intelligent detection system can be guaranteed.
  • the detection method of the above-mentioned intelligent detection system of the present invention is implemented, for example, by a software module that is solidified in a central processing unit or a memory of the intelligent detection system. Alternatively, it can also be implemented as a physical hardware chip, which is solidified in the central processing unit.
  • the program controls the working mode of the intelligent detection system, the working sequence, and the operation of the external hardware chip, and monitors the intelligent detection system itself. Working status.
  • the information collected by each sensor is stored in the memory of the central processing board, and the information can be transmitted to other devices through the RS232 serial port, the 485 bus interface, the network interface, the CAN bus interface, and the photoelectric conversion interface.
  • the invention aims to protect an embedded intelligent detection system and a detection method.
  • the intelligent detection system adopts an embedded system structure and a software programming method, and realizes online data collection of temperature, vibration and acoustic sensors.
  • the invention can effectively replace the existing manual detection means, realize online monitoring and alarm in the operation of the device, and improve the safety of the operation of the device.
  • the failure detection of the mechanical parts of the prior art processing equipment is basically in the manual detection stage, and is concentrated on remote fault diagnosis.
  • the invention realizes the intelligent detection of mechanical failure of the device by transplanting the intelligent processing algorithm and the diagnostic method into the embedded system.

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Human Computer Interaction (AREA)
  • Manufacturing & Machinery (AREA)
  • Automation & Control Theory (AREA)
  • Health & Medical Sciences (AREA)
  • Quality & Reliability (AREA)
  • General Engineering & Computer Science (AREA)
  • Biomedical Technology (AREA)
  • Testing Of Devices, Machine Parts, Or Other Structures Thereof (AREA)
  • Measurement Of Mechanical Vibrations Or Ultrasonic Waves (AREA)

Description

用于设备故障检测的智能检测系统及检测方法 技术领域
本发明涉及一种用于检测设备运行故障的智能检测系统, 结合采用了 智能传感技术和基于嵌入式的计算机科学技术。 具体的, 涉及一种用于设 备故障检测的智能检测系统及检测方法。 背景技术
各种大型设备, 如数控机床、 专用设备等, 在使用过程中, 需要随时 检测设备的运行状况, 以便及时进行维护, 保障设备的安全运行。 在目前 的设备运行状况监测中, 对于设备的电子部分故障和运行状态的监测技术 已经比较完善, 并且能够使用电子设备本身自检的方式排除一些软件和硬 件故障。 但是, 设备机械部分的运行情况和参数的监测一直是故障检测和 诊断领域的难题。 例如对数控机床主轴、 驱动电机、 床身以及刀具库的运 行情况检测, 是早期发现故障隐患, 保证机床安全运行的必要手段, 也是 机床维修、 维护的基本依据。 通常, 旋转部件的温度、 紧固部件的振动和 加工过程中的声学信号变化能够准确的反映机械设备的运行情况。
但是, 现有技术中缺乏有效的技术手段以实时、 准确的对这些状态参 数进行监测, 多数情况下是依赖维修人员的眼睛和耳朵分辨设备运行中的 声音和振动,是否能够准确有效地判断故障点全凭检修人员的经验。因此, 缺乏有效的设备和检测方法以保证设备故障检测的效率和精度。 发明内容
为了解决现有设备机械部分运行情况和参数的检测问题, 本发明的目 的是提供一种用于设备故障检测的智能检测系统及检测方法, 其基于嵌入 式技术实现对大型设备的机械故障进行实时、 智能的检测和诊断。
根据本发明的一个方面, 提供了一种用于设备故障检测的智能检测系 统, 所述智能检测系统外接有用于采集反映待检测设备运行状态的数据的 多个传感器, 该智能检测系统包括: 中央处理板 2, 其具有中央处理单元 以及与该中央处理单元连接的多个数据接口; 数据采集板 3, 其连接到所 述多个传感器的一个或多个, 用于对所述传感器采集的数据进行处理; 同 步通信板 4, 用于使所述中央处理板 2与所述数据采集板 3之间的通信保 持同步; 多个连接插件 1, 通过所述多个连接插件 1连接所述中央处理板 2、 数据采集板 3和同步通信板 4以实现数据传输; 其中, 所述中央处理 单元用于对所述多个传感器采集的数据进行分析, 当采集的数据超过设定 限度时, 表示待检测设备状态异常, 发出报警信息; 当采集的数据未超过 设定限度时, 表示待检测设备状态正常。
优选的,所述多个传感器包括振动传感器、声学传感器和温度传感器。 可选的, 所述中央处理单元对所述温度传感器采集的温度数据进行分 析, 当该温度数据超过预设的上下限时, 发出报警信息。
可选的, 所述中央处理单元对所述振动传感器采集的振动信号进行分 析以获取振动幅度和振动频率特性; 并且预先设定振动幅度的上下限, 当 采集得到的振动幅度超过所设定的上下限时进行报警; 预先设定振动信号 在正常状况下的频率特性, 将采集得到的振动频率特性与该频率特性进行 比较, 当出现异常时发出报警信息。
可选的, 所述中央处理单元对所述声学传感器采集的声学信号进行分 析以获取声压级数值和声音频率特性; 并且预先设定最大声压级数值, 当 采集得到的声压级数值超过该最大声压级数值时发出报警信息; 预先设定 声学信号在正常状况下的频率特性, 将采集得到的声学频率特性与该频率 特性进行比较, 当出现异常时发出报警信息。
可选的, 所述多个连接插件 1采用 PC104总线结构。
可选的,所述多个数据接口包括串口、 485总线接口、 CAN总线接口、 网络接口和光电转换接口。
优选的, 所述温度传感器连接到所述 485总线接口上, 所述振动传感 器和声学传感器连接到所述数据采集板 3上。
优选的, 所述数据采集板 3是多通道高速数据采集板。
可选的, 所述智能检测系统还包括自身工作状态监测模块, 用于监测 该智能检测系统是否存在故障, 当出现故障时出报警信息。 优选的, 所述被监测的智能检测系统自身的工作状态包括: 所述多个 传感器没有釆集到数据, 所述数据采集板 3工作异常, 所述多个数据接口 连接中断。
根据本发明的另一个方面, 提供了一种用于智能检测系统的智能检测 方法, 所述智能检测系统连接有用于采集待检测设备的运行状态数据的多 个传感器, 该智能检测系统包括中央处理板 2、 数据采集板 3、 同步通信 板 4和多个连接插件 1, 通过所述多个连接插件 1连接所述中央处理板 2、 数据采集板 3和同步通信板 4以实现数据传输, 所述智能检测方法包括如 下步骤: 系统初始化步骤, 初始化所述智能检测系统的各个设备和接口, 使其进入就绪状态; 工作方式设定步骤, 将所述智能检测系统的工作方式 设定为主控制器或从控制器; 工作时序设定步骤, 设定所述智能检测系统 的各个设备和接口的工作时序和任务优先级; 数据采集步骤, 所述智能检 测系统按照设定好的工作时序和任务优先级开始执行数据采集任务; 数据 分析步骤, 所述智能检测系统对采集到的数据进行分析, 并根据数据分析 结果判断所述待检测设备是否处于正常状态。
其中, 所述工作时序设定步骤包括设定所述多个传感器的数据采集、 数据处理以及数据传输的工作时序和任务优先级。
优选的, 网络接口的数据传输任务优先级最高, 温度传感器的数据采 集任务优先级次高, 振动传感器和声学传感器的数据采集任务优先级最 其中, 所述数据分析步骤还包括如下步骤: 当采集的数据超过设定限 度时, 表示待检测设备状态异常, 发出报警信息; 当采集的数据未超过设 定限度时, 表示待检测设备状态正常, 返回到前述的工作方式设定步骤, 继续进行下一轮的检测。
其中, 预先设定温度上下限, 当温度传感器采集的温度数据超过设定 的上下限时, 发出报警信息。
其中, 所述数据分析步骤对振动传感器釆集得到的振动信号进行分析 以获取振动幅度和振动频率特性; 并且预先设定振动幅度的上下限, 当采 集得到的振动幅度超过所设定的上下限时进行报警; 预先设定振动信号在 正常状况下的频率特性, 将采集得到的振动频率特性与该频率特性进行比 较, 当出现异常时发出报警信息。
其中, 所述数据分析步骤对声学传感器采集得到的声学信号进行分析 以获取声压级数值和声音频率特性; 并且预先设定最大声压级数值, 当采 集得到的声压级数值超过该最大声压级数值时进行报警; 预先设定声学信 号在正常状况下的频率特性, 将采集得到的声学频率特性与该频率特性进 行比较, 当出现异常时发出报警信息。
可选的, 所述智能检测系统还包括自身工作状态监测步骤, 用于监测 该智能检测系统自身是否存在故障, 当出现故障时出报警信息。 并且, 所 述被监测的智能检测系统自身的工作状态包括: 所述多个传感器没有采集 到数据, 所述数据采集板 3工作异常, 与所述中央处理板 2连接的多个数 据接口连接中断。
如上所述, 根据本发明的智能检测系统, 将温度传感器、 振动传感器 和声学传感器安装在设备的检测部位, 在设备运行中实时检测旋转部件的 温度、 紧固部件的振动和加工过程中的声学信号变化, 从而能够及时发现 故障隐患或故障点, 及时报警维修, 避免了设备损坏。 因此, 大大提高了 设备故障检测的效率和精度, 为设备安全运行提供了有效的保障。 附图说明
图 1是根据本发明实施方式的智能检测系统的结构示意图; 图 2是图 1中所示中央处理板的结构示意图;
图 3显示了本发明的智能检测系统的外壳;
图 4显示了根据本发明的智能检测系统的智能检测方法的流程图; 图 5显示了本发明实施例的 BP神经网络的拓扑结构图;
图 6显示了待检测设备在正常状态下的振动频率特性;
图 7显示了待检测设备在故障状态下的振动频率特性。 具体实施方式
为使本发明的目的、 技术方案和优点更加清楚明了, 下面结合具体实 施方式并参照附图, 对本发明进一步详细说明。 应指出的是, 所描述的实 施例旨在便于对本发明的理解, 而对其不起任何限定作用。 图 1是根据本发明实施方式的智能检测系统的结构示意图。
如图 1所示, 根据本发明的智能检测系统包括中央处理板 2, 数据采 集板 3和同步通信板 4。
中央处理板 2的数据总线采用了兼容 PC104总线的结构,该数据总线 通常采用排状的接插件实现, 接插件的一端是插孔, 另一端是带插针的连 接插件 1。 本发明的中央处理板 2, 数据采集板 3和同步通信板 4均带有 这样的连接插件 1, 从而使得各电路板能够连接在一起以便于传输数据。 在本发明的实施例中, 中央处理板 2优选的采用嵌入式中央处理板, 其上 搭载有中央处理单元, 用于对整个智能检测系统的运行进行控制。
同步通信板 4连接在中央处理板 2与数据采集板 3之间, 其上设置有 时序调整电路, 该同步通信板 4通过对所述中央处理板 2与所述数据采集 板 3之间的通信进行时序调整, 使得二者之间的通信保持同步。 同步通信 板 4上设置有连接插件 1, 并通过连接插件 1与中央处理板 2和数据采集 板 3实现电气连接。
数据采集板 3包括多个数据采集通道和寄存器, 寄存器用于存储当前 时刻采集通道的数据, 中央处理单元通过 PC104 总线读取寄存器中的数 据。 本发明中的数据采集板 3优选的采用多通道高速数据采集板, 例如是 DM6430HR- l o
图 2是图 1中所示中央处理板的结构示意图。
如图 2所示,根据本发明实施例的中央处理板主要包括中央处理单元, 该中央处理单元优选的采用 ARM9型号的微处理器芯片。本发明的中央处 理单元可以扩展存储器, 用于存储程序运行文件和临时数据。 本发明中, 存储器优选的为 FLASH存储器 (即闪存), 以保证较高的数据读取速率, 可选的, 该存储器为 32M容量。
本发明的智能检测系统采用 PC104总线结构,中央处理单元能够通过 PC104总线与数据采集板 3、 同步通信板 4等外部设备连接, 并控制这些 外部设备工作。
如图 2所示, 中央处理板上还集成有多个数据接口, 这些数据接口连 接在外部设备和中央处理单元之间, 用作数据传输和指令的通讯接口。 这 些数据接口分别对应于不同的数据接口标准, 例如串口、 485总线接口、 CAN总线接口(CAN总线是一种总线标准)、网络接口、光电转换接口等。 这里, 所述串口优选的是标准的 232串口。
本发明中通过多个传感器来采集反映待检测设备实时运行状态的状 态数据, 例如振动传感器、 声学传感器、 温度传感器、 压力传感器、 加速 度传感器等。 不同的传感器输出数据类型不同, 因而需要用不同的数据传 输接口把这些数据分别传输到中央处理板 2上的中央处理单元。在本发明 中, 根据传感器输出数据的不同类型, 不同的传感器连接到不同标准的接 口, S卩 232接口、 485接口、 CAN接口, 网络接口和光电转换接口等。 在本发明的实施例中, 温度传感器连接到 485接口上, 而振动传感器和声 学传感器直接连接到数据采集板 3上, 用于采集反映待检测设备实时运行 状态的状态参数, 如振动、 噪音等。 例如, 作为优选实施方式之一, 使用 485接口连接了 16个温度传感器, 而使用高速数据采集板 HR6430-1连接 了 Ί个振动传感器和 4个声学传感器。
另外, 本发明的智能检测系统配备有电源以及对应的供电电路, 用于 为中央处理单元、数据采集板和外部连接的传感器提供恒定电压,例如 3V, 5V, 12V的恒定电压。
图 3显示了本发明的智能检测系统的外壳。
如图 3所示, 本发明的智能检测系统封装在外壳 3-1 中, 该外壳 3-1 用于容纳并保护设置在其内部的智能检测系统的组成部件。
如图 3所示, 外壳 3-1的两个侧面分别设置有开口, 并安装有插座, 供电源、 传感器引线和外部通讯接口使用。
外壳 3-1的正面设置有 2个开口, 分别安装有 2个 D型插座(一种通 用插座, 外形如英文字母 D), 一个是电源插座 3-2, 可以连接 220v电源, 另一个是传感器引线插头 3-3。
外壳 3-1的一个侧面设置有 4个开口, 分别连接 RS232接口 3-4、 485 接口 3-5、 CAN总线接口 3-6和网络接口 3-7, 这些接口用于连接外部线 路, 实现数据信息的传输。
图 4显示了根据本发明的智能检测系统的智能检测方法的流程图。 如图 4所示, 本发明的智能检测方法包括如下步骤:
系统初始化步骤。 即初始化智能检测系统的各个设备和接口, 使系统 进入就绪状态。 具体来说, 对中央处理板 2、 数据采集板 3、 同步通信板 4 进行检测, 并确定通过外部接口连接的设备和设备状况是否正常。
工作方式设定步骤。 根据待检测设备的设定, 将智能检测系统的工作 方式设定为主控制器或者从控制器。 例如, 如果待检测设备设定为希望本 发明的智能检测系统独立工作, 则将该智能检测系统的工作方式设定为主 控制器。 此时, 由智能检测系统产生各种控制命令, 指挥其他外部设备工 作。 另一方面, 如果待检测设备设定为希望控制本发明的智能检测系统, 则将该智能检测系统的工作方式设定为从控制器。 此时, 智能检测系统作 为外部主控制器的外接设备, 响应外部主控制器发送的各种控制命令。 在 本实例中, 以设定智能检测系统作为主控制器为例给出工作流程。 此外, 工作方式的设定可以按照预先设定自动进行, 也可以在智能检测系统的运 行过程中, 通过操作员人工设定。
工作时序设定步骤。 根据待检测设备的类型, 设定智能检测系统的各 个设备和接口的工作时序和任务优先级,使其保持正常的工作状态。例如, 如果待检测设备对温度参数敏感, 则设定温度传感器的数据采集优先级 高; 如果对振动或噪声敏感, 则设定振动传感器或声学传感器的数据采集 优先级高。所述工作时序设定步骤例如包括设定所述多个传感器的数据采 集、 数据处理以及数据传输的工作时序和任务优先级等。
作为在本发明的一个优选实施方式, 例如可以设定如下: 网络接口的 数据传输任务优先级最高, 通过 485总线连接的温度传感器采样任务优先 级次高, 通过数据采集板连接的振动传感器和声学传感器的数据采集任务 优先级最低。
数据采集步骤。 智能检测系统按照设定好的工作时序和任务优先级, 开始执行数据采集任务。 即通过设置在待检测设备的各个检测位置上的传 感器采集反映待检测设备运行状态的数据。
数据分析步骤。 智能检测系统对采集到的数据进行分析, 并根据数据 分析结果判断待检测设备是否处于正常状态。 具体来说, 当采集的数据超 过设定限度时, 表述待检测设备状态异常, 发出报警信息; 当采集的数据 未超过设定限度时, 表述待检测设备状态正常, 返回到前面的工作方式设 定步骤, 继续进行下一轮的检测。 本发明中, 具体的是温度传感器、 振动传感器和声学传感器, 针对各 传感器采集数据的分析方法如下。
温度数据: 设定温度上下限, 当温度传感器采集的温度数值超过设定 的上下限时, 发出报警信息。
振动数据: 振动传感器采集对象设备的振动信号, 当该振动信号异常 时, 发出报警信息。 对振动信号的分析包括振动幅度和 /或振动频率。
对于振动幅度, 可以预先设定振动幅度的上下限。 分析时, 通过对振 动信号进行时域处理以获得振幅, 当振幅超过所设定的振幅上下限时进行 报警。
对于振动频率, 可以预先设定振动信号在正常状况下的频率特性。 分 析时, 采用傅里叶变换算法或小波分析算法, 将振动信号转换为周期性的 频域值, 与事先设定好的正常状况下振动信号的频率特性进行比较, 当出 现异常时发出报警信息。
声学数据: 声学信号的处理与振动信号类似, 设定最大声压级数值和 正常状况下声学信号的频率特性, 通过对声学信号进行时域处理获得声压 级数值, 当声压级数值超过设定的最大声压级数值时进行报警。 另外, 通 过傅里叶变换算法或小波分析算法得到声学信号的频率特性, 与预先设定 好的正常状况下声学信号的频率特性进行比较, 当出现异常时发出报警信 息。
下面, 将以数控机床主轴的机械故障为例, 介绍本发明的智能检测系 统的一个具体实施例。
为检测数控机床主轴的故障情况, 可以按如下方式设置传感器。 温度传感器: 2个温度传感器安装在主轴的轴套上, 温度传感器的上 限设定为 80摄氏度, 下限设定为 10摄氏度。
检测时, 如果温度传感器采集的温度值在 10-80摄氏度之间, 则表示 设备状态正常; 否则, 如果超出该数值范围的上下限, 则发出报警信息。
振动传感器: 2个振动传感器分别安装在主轴的轴向和径向上, 用来 检测主轴在轴向和径向的振动, 主轴径向和轴向的振动振幅的上下限分别 设定为 ±0.5mm。
检测时, 如果振动传感器采集的振动振幅数值在 ± 0.5mm之间, 则表 示设备状态正常; 否则, 如果超出该数值范围的上下限, 则发出报警信息。 声学传感器: 4个声学传感器分别安装在主轴的周围, 探头对着主轴 方向, 用来收集主轴发出的声音, 主轴的声压级上限设定为 94DB, 下限 设定为 50DB。
检测时, 如果声学传感器采集的声压级数值在 50DB至 94DB之间, 则表示设备状态正常; 否则, 如果超出该数值范围的上下限, 则发出报警 信息。
下面以振动信号为例, 详细介绍振动信号的频率特性的采集和处理过 程。
首先介绍振动信号在正常状况下的频率特性的获取。
首先, 振动传感器采集主轴正常运行时的振动信号 (表现为振动波形 的形式), 采样频率为 10kHz, 采样输出为 ± 5v。 其次, 对每一次采样获 得的振动波形, 利用 Daubenchies小波系列的 4 阶小波进行 3层小波包分 解, 得到 8个分量的频带。 再次, 对该 8个频带的信号分别求取总能量, 以能量为元素构造特征向量, 特征向量经过归一化处理, 得到归一化后的 特征向量。单次采样所获得的振动信号的归一化特征向量如下面的表 1所 示:
1
Figure imgf000011_0001
按照同样的采样和处理方式, 对上述过程重复 14次, 得到 14组主轴 正常运行时的振动信号, 经上述处理后进而得到 14组归一化特征向量, 如上面的表 2所示。
随后, 以前述步骤得到的 14组特征向量作为输入, 0为输出, 对系统 预先设置的一神经网络进行训练。
图 5显示了本发明实施例的 B P神经网络的拓扑结构图。
如图 5 所示, 本发明的智能检测系统中预先建立一个 BP ( back propagation)神经网络(误差反向传递神经网络),该神经网络包括输入层、 中间层和输出层。 其中, 输入层用于输入实时采集信号经过小波变化得到 的归一化后特征向量, 包括 8个神经元节点, 每个神经元节点对应输入每 组特征向量的 8个分量中的一个; 中间层包括 4个神经元节点, 用于对于 输入层输入的数据进行处理, 以提高神经网络的计算精度; 输出层包括 2 个神经元节点, 输出层的输出值为 0或 1 , 分别表示状态正常或状态异常 (即存在故障)。
本实施例中, 中间层、 输出层的节点数量与输入层的节点数量之间没 有必然联系, 可以根据实际情况确定节点数量。 另外, 中间层神经元选择 tansig ( S型正切函数)为激活函数, 输出层神经元选择 logsig (S 型对数函 数) 为激活函数。 但是, 本领域技术人员应该知道, 神经网络的结构、 神 经元的数量、 神经元的激活函数根据实际情况可以改变。
本发明的神经网络的设计要求是: 神经网络训练最多迭代 20000次, 输出误差小于 0.002。
对该神经网络训练好之后, 即可得到机床主轴在正常状态下的振动频 率特性, 如图 6所示。
图 6显示了待检测设备在正常状态下的振动频率特性。
如图 6所示, 显示了正常状态下每组振动信号的 8个分量的波形, 在 通过图 6右边所示的神经网络训练之后, 输出故障分类的分类号为 0, 即 表示状态正常。
在实时的故障检测过程中, 振动传感器实时采集机床主轴的振动信 号, 并按照上述步骤对该信号进行处理。 当不正常的振动信号经过小波变 换得到的归一化特征向量输入上述神经网络后, 神经网络的输出为 1, 即 表示存在故障, 系统发出报警信息。
图 7显示了待检测设备在故障状态下的振动频率特性。
如图 7所示, 显示了故障状态下每组振动信号的 8个分量的波形。 通 过与图 6的对比发现, 故障状态下每组振动信号的 8个分量的波形与正常 状态下对应分量的波形存在明显差异。 进一步, 在通过图右边的神经网络 运算后, 输出故障分类的分类号为 1, 即表示存在故障, 系统会发出报警 信息。
本领域技术人员应该知道, 本实施例中的重复次数 14不是固定的, 而是根据实际情况可多可少。 本实例中, 重复 14次即可达到设计要求: 神经网络训练最多迭代 20000次, 输出误差小于 0.002。
声学信号的处理过程与上述对振动信号的处理过程相同。 声学信号同 样经过小波变换, 得到归一化特征向量, 使用正常状态下的特征向量训练 神经网络, 得到声学信号在正常状态下的频率特性。 当异常的声音信号 输入网络时, 网络输出为 1, 即表示发现故障, 系统发出报警信息。
另外, 在智能检测系统设置为从控制器的情况下, 工作流程与上述主 控制器的设定步骤类似, 区别仅在于工作时序设定由外部主控制器执行。
可选的, 除了上述流程以外, 本发明的智能检测系统还设置有自身工 作状态监测模块 (未显示), 用于监测智能检测系统自身的工作状态是否 正常。 并且, 该自身工作状态监测模块的工作流程与上述检测待检测设备 工作状态的流程相互独立的运行。本发明的智能检测系统的故障状态或非 正常状态例如包括如下情况: 通过各数据接口 (如 RS232接口、 485总线 接口)接入智能检测系统的传感器没有数据,通过 PC104总线连接到中央 处理单元的数据采集板工作异常, 网络通讯连接中断等。 当出现上述故障 状态时, 智能检测系统及时发出报警信息。 这样, 可以保证智能检测系统 可靠的运行和及时的维护。
本发明的上述智能检测系统的检测方法例如由固化在智能检测系统 的中央处理单元或存储器中的软件模块实现, 可选的, 也可以实现为实体 的硬件芯片, 固化在中央处理单元中的主程序控制着智能检测系统的工作 方式、 工作时序, 以及外接的硬件芯片的运行, 并监测智能检测系统自身 的工作状态。此外,各传感器采集到的信息存储在中央处理板的存储器中, 并且能够通过 RS232串口, 485总线接口、 网络接口、 CAN总线接口、 光 电转换接口将信息传输到其他设备中。
本发明旨在保护一种嵌入式智能检测系统及检测方法, 该智能检测系 统采用嵌入式系统结构和软件编程方法, 实现了温度、 振动以及声学传感 器的在线数据采集。 本发明能够有效替代现有的人工检测手段, 在设备运 行中实现在线监测和报警, 提高了设备运行的安全性。
如上所述, 现有技术的加工设备 (如数控机床等) 的机械部分的故障 检测基本上处在人工检测阶段, 并集中在远程故障诊断上。 而本发明通过 将智能处理算法和诊断方法移植到嵌入式系统中, 实现了设备机械故障的 智能检测。
应当理解的是, 本发明的上述具体实施方式仅仅用于示例性说明或解 释本发明的原理, 而不构成对本发明的限制。 因此, 在不偏离本发明的精 神和范围的情况下所做的任何修改、 等同替换、 改进等, 均应包含在本发 明的保护范围之内。 此外, 本发明所附权利要求旨在涵盖落入所附权利要 求范围和边界、 或者这种范围和边界的等同形式内的全部变化和修改例。

Claims

权利 要求 书
1、 一种用于设备故障检测的智能检测系统, 所述智能检测系统外接 有用于采集反映待检测设备运行状态的数据的多个传感器, 该智能检测系 统包括:
中央处理板(2), 其具有中央处理单元以及与该中央处理单元连接的 多个数据接口;
数据采集板(3 ), 其连接到所述多个传感器的一个或多个, 用于对所 述传感器采集的数据进行处理;
同步通信板(4), 用于使所述中央处理板(2)与所述数据采集板(3 ) 之间的通信保持同步;
多个连接插件 (1 ), 通过所述多个连接插件 (1 ) 连接所述中央处理 板 (2)、 数据釆集板 (3 ) 和同步通信板 (4) 以实现数据传输;
其特征在于- 所述中央处理单元用于对所述多个传感器采集的数据进行分析, 当采 集的数据超过设定限度时, 表示待检测设备状态异常, 发出报警信息; 当采集的数据未超过设定限度时, 表示待检测设备状态正常。
2、 根据权利要求 1 所述的智能检测系统, 其中, 所述多个传感器包 括振动传感器、 声学传感器和温度传感器。
3、 根据权利要求 2所述的智能检测系统, 其中, 所述中央处理单元 对所述温度传感器采集的温度数据进行分析, 当该温度数据超过预设的上 下限时, 发出报警信息。
4、 根据权利要求 2所述的智能检测系统, 其中, 所述中央处理单元 对所述振动传感器采集的振动信号进行分析以获取振动幅度和振动频率 特性; 并且
预先设定振动幅度的上下限, 当采集得到的振动幅度超过所设定的上 下限时进行报警;
预先设定振动信号在正常状况下的频率特性, 将采集得到的振动频率 特性与该频率特性进行比较, 当出现异常时发出报警信息。
5、 根据权利要求 2所述的智能检测系统, 其中, 所述中央处理单元 对所述声学传感器釆集的声学信号进行分析以获取声压级数值和声音频 率特性; 并且
预先设定最大声压级数值, 当采集得到的声压级数值超过该最大声压 级数值时发出报警信息;
预先设定声学信号在正常状况下的频率特性, 将采集得到的声学频率 特性与该频率特性进行比较, 当出现异常时发出报警信息。
6、 根据权利要求 1 所述的智能检测系统, 其中, 所述多个连接插件 ( 1 ) 采用 PC104总线结构。
7、根据权利要求 1所述的智能检测系统, 其中, 所述中央处理板(2) 还包括存储器, 用于存储程序运行文件和临时数据; 并且
所述中央处理单元是基于嵌入式的微处理芯片;
多个数据接口用于与外部设备连接以传输数据和指令。
8、 根据权利要求 1 所述的智能检测系统, 其中, 所述多个数据接口 包括串口、 485总线接口、 CAN总线接口、 网络接口和 /或光电转换接口。
9、 根据权利要求 8所述的智能检测系统, 其中, 每个所述多个传感 器根据其输出数据的不同类型分别连接到相应接口标准的数据接口。
10、 根据权利要求 9所述的智能检测系统, 其中, 所述温度传感器连 接到所述 485总线接口上, 所述振动传感器和声学传感器连接到所述数据 采集板 (3 ) 上。
11、根据权利要求 1-10中任一项所述的智能检测系统, 其中, 所述数 据采集板 (3 ) 是多通道高速数据采集板。
12、根据权利要求 1-10中任一项所述的智能检测系统, 其中, 所述智 能检测系统还包括电源以及对应的供电电路,用于为所述中央处理板(2)、 数据采集板 (3 ) 和外接的所述多个传感器提供恒定电压。
13、根据权利要求 1-10中任一项所述的智能检测系统, 其中, 所述智 能检测系统还具有一外壳 (3-1 ), 其用于容纳并保护所述智能检测系统的 组成部件。
14、 根据权利要求 13所述的智能检测系统, 其中, 所述外壳 (3-1 ) 的两个侧面分别设置有开口, 用于安装供电源、 传感器引线和外部通讯接 口使用的插座。
15、根据权利要求 1-10中任一项所述的智能检测系统, 其中, 所述智 能检测系统还包括自身工作状态监测模块, 用于监测该智能检测系统是否 存在故障, 当出现故障时出报警信息。
16、 根据权利要求 15所述的智能检测系统, 其中, 所述被监测的智 能检测系统自身的工作状态包括: 所述多个传感器没有采集到数据, 所述 数据采集板 (3 ) 工作异常, 所述多个数据接口连接中断。
17、 一种用于智能检测系统的智能检测方法, 所述智能检测系统连接 有用于采集待检测设备的运行状态数据的多个传感器, 该智能检测系统包 括中央处理板 (2)、 数据采集板 (3 )、 同步通信板 (4 ) 和多个连接插件
( 1 ), 通过所述多个连接插件 (1 ) 连接所述中央处理板 (2)、 数据采集 板 (3 ) 和同步通信板 (4) 以实现数据传输, 所述智能检测方法包括如下 步骤:
系统初始化步骤, 初始化所述智能检测系统的各个设备和接口, 使其 进入就绪状态;
工作方式设定步骤, 将所述智能检测系统的工作方式设定为主控制器 或从控制器;
工作时序设定步骤, 设定所述智能检测系统的各个设备和接口的工作 时序和任务优先级; 数据采集步骤, 所述智能检测系统按照设定好的工作时序和任务优先 级开始执行数据采集任务;
数据分析步骤, 所述智能检测系统对采集到的数据进行分析, 并根据 数据分析结果判断所述待检测设备是否处于正常状态。
18、 根据权利要求 17所述的方法, 其中, 所述系统初始化步骤包括 对所述中央处理板 (2)、 数据采集板 (3 )、 同步通信板 (4) 进行检测, 并确定通过外部接口连接的设备和设备状况是否正常。
19、 根据权利要求 17所述的方法, 其中, 所述工作方式设定步骤还 包括如下步骤:
当所述智能检测系统设定为主控制器时, 由该智能检测系统产生各种 控制命令以指挥其他外部设备工作;
当所述智能检测系统设定为从控制器时, 由该智能检测系统作为外部 主控制器的外接设备, 响应外部主控制器发送的各种控制命令。
20、 根据权利要求 17所述的方法, 其中, 所述工作时序设定步骤包 括设定所述多个传感器的数据采集、数据处理以及数据传输的工作时序和 任务优先级。
21、 根据权利要求 17-20中任一项所述的方法, 其中, 所述数据分析 步骤还包括如下步骤:
当采集的数据超过设定限度时, 表示待检测设备状态异常, 发出报警 信息;
当采集的数据未超过设定限度时, 表示待检测设备状态正常, 返回到 前述的工作方式设定步骤, 继续进行下一轮的检测。
22、 根据权利要求 21 所述的方法, 其中, 预先设定温度上下限, 当 温度传感器采集的温度数据超过设定的上下限时, 发出报警信息。
23、 根据权利要求 21 所述的方法, 其中, 所述数据分析步骤对振动 传感器采集的振动信号进行分析以获取振动幅度和振动频率特性; 并且 预先设定振动幅度的上下限, 当采集得到的振动幅度超过所设定的上 下限时进行报警;
预先设定振动信号在正常状况下的频率特性, 将采集得到的振动频率 特性与该频率特性进行比较, 当出现异常时发出报警信息。
24、 根据权利要求 21 所述的方法, 其中, 所述数据分析步骤对声学 传感器采集的声学信号进行分析以获取声压级数值和声音频率特性; 并且 预先设定最大声压级数值, 当采集得到的声压级数值超过该最大声压 级数值时发出报警信息;
预先设定声学信号在正常状况下的频率特性, 将采集得到的声学频率 特性与该频率特性进行比较, 当出现异常时发出报警信息。
25、 根据权利要求 17-20中任一项所述的方法, 其中, 所述智能检测 系统还包括自身工作状态监测步骤, 用于监测该智能检测系统自身是否存 在故障, 当出现故障时出报警信息。
26、 根据权利要求 25所述的方法, 其中, 所述被监测的智能检测系 统自身的工作状态包括: 所述多个传感器没有采集到数据, 所述数据釆集 板(3 )工作异常, 与所述中央处理板(2)连接的多个数据接口连接中断。
PCT/CN2010/080579 2010-12-31 2010-12-31 用于设备故障检测的智能检测系统及检测方法 WO2012088707A1 (zh)

Priority Applications (3)

Application Number Priority Date Filing Date Title
US13/976,882 US20140298099A1 (en) 2010-12-31 2010-12-31 Intelligent detection system and method for detecting device fault
PCT/CN2010/080579 WO2012088707A1 (zh) 2010-12-31 2010-12-31 用于设备故障检测的智能检测系统及检测方法
CN2010800706166A CN103250107A (zh) 2010-12-31 2010-12-31 用于设备故障检测的智能检测系统及检测方法

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
PCT/CN2010/080579 WO2012088707A1 (zh) 2010-12-31 2010-12-31 用于设备故障检测的智能检测系统及检测方法

Publications (1)

Publication Number Publication Date
WO2012088707A1 true WO2012088707A1 (zh) 2012-07-05

Family

ID=46382224

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2010/080579 WO2012088707A1 (zh) 2010-12-31 2010-12-31 用于设备故障检测的智能检测系统及检测方法

Country Status (3)

Country Link
US (1) US20140298099A1 (zh)
CN (1) CN103250107A (zh)
WO (1) WO2012088707A1 (zh)

Cited By (11)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105320071A (zh) * 2014-05-26 2016-02-10 发那科株式会社 数值控制装置
EP2926286A4 (en) * 2012-11-30 2016-08-03 Discovery Sound Technology LLC SYSTEM AND METHOD FOR MONITORING EQUIPMENT SOUND
US9971667B1 (en) 2012-11-30 2018-05-15 Discovery Sound Technology, Llc Equipment sound monitoring system and method
US10145761B1 (en) 2012-11-30 2018-12-04 Discovery Sound Technology, Llc Internal arrangement and mount of sound collecting sensors in equipment sound monitoring system
US10156844B1 (en) 2012-11-30 2018-12-18 Discovery Sound Technology, Llc System and method for new equipment configuration and sound monitoring
CN111948966A (zh) * 2020-08-14 2020-11-17 江苏师范大学 一种基于物联网的防过卷无线收发模块
CN111982200A (zh) * 2020-08-26 2020-11-24 台州职业技术学院 一种机电设备故障在线预警检测设备及其检测方法
CN112147647A (zh) * 2020-08-11 2020-12-29 徐州徐工挖掘机械有限公司 一种监控预测gps终端故障的方法及系统
US11188292B1 (en) 2019-04-03 2021-11-30 Discovery Sound Technology, Llc System and method for customized heterodyning of collected sounds from electromechanical equipment
CN116021425A (zh) * 2023-03-28 2023-04-28 江苏中科云控智能工业装备有限公司 一种基于物联网的去毛刺设备性能测试系统及方法
US11988547B2 (en) 2021-01-04 2024-05-21 Central South University Train compartment vibration monitoring method and vibration signal feature library establishment and application methods

Families Citing this family (16)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
GB2500388B (en) * 2012-03-19 2019-07-31 Ge Aviat Systems Ltd System monitoring
US9558053B1 (en) * 2015-06-26 2017-01-31 Amazon Technologies, Inc. Computer systems monitoring using beat frequency analysis
DE102016008987B4 (de) * 2015-07-31 2021-09-16 Fanuc Corporation Maschinenlernverfahren und Maschinenlernvorrichtung zum Lernen von Fehlerbedingungen, und Fehlervorhersagevorrichtung und Fehlervorhersagesystem, das die Maschinenlernvorrichtung einschließt
CN106886499A (zh) * 2015-12-15 2017-06-23 西安富成防务科技有限公司 主被动传感器多模信息探测与信息融合技术仿真平台
WO2017167656A1 (de) * 2016-03-29 2017-10-05 Voith Patent Gmbh Verfahren zur fernüberwachung einer industriellen anlage
US20200173887A1 (en) * 2017-05-25 2020-06-04 Nec Corporation Processing apparatus, processing method, and non-transitory storage medium
EP3425460A1 (de) * 2017-07-04 2019-01-09 Siemens Aktiengesellschaft Vorrichtung und verfahren zur bestimmung des zustands einer spindel einer werkzeugmaschine
CN110718231A (zh) * 2019-09-12 2020-01-21 深圳市铭华航电工艺技术有限公司 基于声学网络的监测方法、装置、终端及存储介质
CN110994795A (zh) * 2019-12-06 2020-04-10 国网江苏省电力有限公司南通供电分公司 一种远程信号采集系统及其采集方法
CN110996289A (zh) * 2019-12-11 2020-04-10 湃方科技(天津)有限责任公司 一种用于机器状态监测诊断仪的智能网关、方法及系统
CN111930081A (zh) * 2020-07-15 2020-11-13 西门子工厂自动化工程有限公司 用于监控agv状态的方法、装置、边缘设备和存储介质
CN112230603B (zh) * 2020-10-14 2021-08-13 深圳吉兰丁智能科技有限公司 一种基于数控机床的多传感器数据采集方法及系统
CN113284327A (zh) * 2021-06-21 2021-08-20 中国特种设备检测研究院 一种适用于过山车振动监测的方法与系统
CN115079639B (zh) * 2022-07-13 2022-11-18 深圳市协和传动器材有限公司 凸轮分割器的加工机床运行异常报警方法
CN117346842B (zh) * 2023-11-14 2024-07-12 广东材通实业有限公司 一种电机接线盒的故障检测方法及系统
CN117268535B (zh) * 2023-11-22 2024-01-26 四川中测仪器科技有限公司 一种基于振动数据的电机转轴状态监测方法

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
KR20020050937A (ko) * 2000-12-22 2002-06-28 김형벽ㅂ 전동차 고장상태 감시 장치
CN1593849A (zh) * 2004-07-15 2005-03-16 无锡开源机床集团有限公司 基于网络的磨床数据采集与故障诊断装置
CN101118437A (zh) * 2007-09-03 2008-02-06 石毅 一种新型的数控机床远程状态监测与故障诊断系统实现方法
CN101324657A (zh) * 2008-07-15 2008-12-17 武汉华瑞测控科技有限公司 变电站高压计量用互感器在线监测方法
CN101556476A (zh) * 2009-05-19 2009-10-14 厦门大学 一种用于规模化机械加工的无线网络检测系统

Family Cites Families (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US7020802B2 (en) * 2002-10-17 2006-03-28 Sun Microsystems, Inc. Method and apparatus for monitoring and recording computer system performance parameters
US8046637B2 (en) * 2009-03-06 2011-10-25 Oracle America, Inc. Telemetry data filtering through sequential analysis
CN101923336B (zh) * 2009-06-09 2012-08-29 鸿富锦精密工业(深圳)有限公司 机台运动控制系统及方法
WO2011002798A2 (en) * 2009-06-30 2011-01-06 Lam Research Corporation Automatic fault detection and classification in a plasma processing system and methods thereof

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
KR20020050937A (ko) * 2000-12-22 2002-06-28 김형벽ㅂ 전동차 고장상태 감시 장치
CN1593849A (zh) * 2004-07-15 2005-03-16 无锡开源机床集团有限公司 基于网络的磨床数据采集与故障诊断装置
CN101118437A (zh) * 2007-09-03 2008-02-06 石毅 一种新型的数控机床远程状态监测与故障诊断系统实现方法
CN101324657A (zh) * 2008-07-15 2008-12-17 武汉华瑞测控科技有限公司 变电站高压计量用互感器在线监测方法
CN101556476A (zh) * 2009-05-19 2009-10-14 厦门大学 一种用于规模化机械加工的无线网络检测系统

Cited By (11)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP2926286A4 (en) * 2012-11-30 2016-08-03 Discovery Sound Technology LLC SYSTEM AND METHOD FOR MONITORING EQUIPMENT SOUND
US9971667B1 (en) 2012-11-30 2018-05-15 Discovery Sound Technology, Llc Equipment sound monitoring system and method
US10145761B1 (en) 2012-11-30 2018-12-04 Discovery Sound Technology, Llc Internal arrangement and mount of sound collecting sensors in equipment sound monitoring system
US10156844B1 (en) 2012-11-30 2018-12-18 Discovery Sound Technology, Llc System and method for new equipment configuration and sound monitoring
CN105320071A (zh) * 2014-05-26 2016-02-10 发那科株式会社 数值控制装置
US11188292B1 (en) 2019-04-03 2021-11-30 Discovery Sound Technology, Llc System and method for customized heterodyning of collected sounds from electromechanical equipment
CN112147647A (zh) * 2020-08-11 2020-12-29 徐州徐工挖掘机械有限公司 一种监控预测gps终端故障的方法及系统
CN111948966A (zh) * 2020-08-14 2020-11-17 江苏师范大学 一种基于物联网的防过卷无线收发模块
CN111982200A (zh) * 2020-08-26 2020-11-24 台州职业技术学院 一种机电设备故障在线预警检测设备及其检测方法
US11988547B2 (en) 2021-01-04 2024-05-21 Central South University Train compartment vibration monitoring method and vibration signal feature library establishment and application methods
CN116021425A (zh) * 2023-03-28 2023-04-28 江苏中科云控智能工业装备有限公司 一种基于物联网的去毛刺设备性能测试系统及方法

Also Published As

Publication number Publication date
US20140298099A1 (en) 2014-10-02
CN103250107A (zh) 2013-08-14

Similar Documents

Publication Publication Date Title
WO2012088707A1 (zh) 用于设备故障检测的智能检测系统及检测方法
CN105573251B (zh) 圆盘式刀库及自动换刀系统综合性能检测平台
CN101579250B (zh) 一种手术电钻的智能控制装置
CN101943682B (zh) 基于阵列式声发射器的分布式机械故障实时监测装置及方法
CN103486068B (zh) 智能风扇的控制检测方法及装置、智能风扇
CN103786071A (zh) 基于无线传感器网络的机床工况实时监控系统
JP2012038302A5 (zh)
CN105527945B (zh) 实现智能健康穿戴设备的快速批量生产、检测与设备信息记录的系统及方法
TW201226101A (en) Method and device to detect the state of cutting tool in machine tool with multiple sensors
CN202880694U (zh) 一种起重机故障诊断装置
CN205620785U (zh) 圆盘式刀库及自动换刀系统综合性能检测平台
TW201925942A (zh) 一種智能診斷系統與方法
CN106154894A (zh) 基于双mpu的整车控制器及其安全监控方法
CN110088462A (zh) 风力涡轮发电机控制系统中的电气故障检测
CN116674388B (zh) 一种基于车载传感器的车辆监测分析系统
CN113432863A (zh) 阀门健康状态检测及诊断装置及方法
CN108361020A (zh) 基于虚拟仪表的坑道钻机用诊断保护装置及方法
CN108919041A (zh) 一种基于聚类分析的变压器绕组状态在线监测方法
CN111722142B (zh) 一种服务器电源信号传输质量检测装置及方法
CN205572058U (zh) 一种用于机床数据采集及故障监测的装置
CN105353645B (zh) 一种电磁阀试验装置及试验方法
CN105300199B (zh) 用于导弹自动测试的rs‑422/485通讯测试监测装置及监测方法
CN209167892U (zh) 一种通用型设备工作参数数据采集卡
CN108037703B (zh) 角位移传感器安装位置的便携式校准装置及校准方法
CN110542475A (zh) 一种数字式振动信号智能变送器

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 10861450

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

WWE Wipo information: entry into national phase

Ref document number: 13976882

Country of ref document: US

122 Ep: pct application non-entry in european phase

Ref document number: 10861450

Country of ref document: EP

Kind code of ref document: A1