KR101057030B1 - WOS-based rotary machine monitoring device and method using DSP transmission method - Google Patents

Abstract

The present invention is to improve the problem that the existing equipment that monitors and analyzes the failure of the rotary machine used in the industrial equipment at a remote location is overloaded when the wireless data transmission to the remote site is difficult to diagnose and monitor the failure at the remote site. The analog signal output from the sensor unit is converted into a digital signal, the converted digital signal is collected for a predetermined time, and the vibration data obtained by FFT conversion is transferred to the sink node unit and the remote server unit through a wireless communication network. The DPCM sensor node unit is configured to transmit, preventing overload of wireless data transmission and providing a GUI map of failure diagnosis of multiple machine facilities at a glance. Maintenance of each facility is facilitated by making timely replacement and repair of And to provide a USN-based rotating machine monitoring device and method using a DPCM transmission method that can prevent accidents caused by emergency trouble.

Sensor section, DPCM sensor node section, fault diagnosis map

Description

WINS-based rotating machine monitoring apparatus and method using DPCM transmission technique {THE APPARATUS AND METHOD OF MONITORING A ROTARY MACHINE WITH DIFFERENT PULSE CODE MODULATION IN UBIQUITOUS SENSOR NETWORK}

The present invention relates to a USN-based rotating machine monitoring apparatus and method using a DPCM transmission technique that can analyze, diagnose, and remotely guide the causes of failure of a plurality of machine facilities in a ubiquitous sensor network environment.

Currently, the system used for the vibration control of mechanical equipment is applied in parallel with the on-line monitoring system for observing the real-time equipment status of some specific equipment and the periodic vibration measurement management method for the majority of the equipment. In particular, they have only been concerned with the high and low values of vibration values and have been dedicated to the ability to quickly communicate facility failures to managers.

Recognizing the importance of the Root Cause Failure Analysis (RCFA) concept in recent years, however, companies are investing heavily in the diagnosis of rotating machinery, and a number of mechanical health managers are being trained each year to diagnose machinery. Diagnosis techniques have evolved over time, but there was no effective program to effectively manage and manage the results of these machines.

In order to solve this problem, Korean Patent Registration No. 10-0850823 describes a number of mechanical facilities; One or more vibration measuring devices for measuring vibrations of the plurality of mechanical facilities; A first terminal receiving and guiding a vibration measurement result and receiving a vibration diagnosis result according to the vibration measurement result guide from an administrator and providing the same to a server; A second terminal configured to receive the vibration diagnosis result and to display a machine facility vibration diagnosis map indicating different states of a plurality of rotating machine facilities according to the vibration diagnosis result; And receiving the vibration measurement result from the vibration measuring device in a database and transmitting the vibration measurement result to the first terminal, and receiving the vibration diagnosis result according to the vibration measurement result for the plurality of machine facilities from the first terminal. Although a machine facility vibration management system and method including a server storing in a database and transmitting to the second terminal have been presented,

It is not a ubiquitous network environment, but consists of a machine equipment vibration management system through wired and wireless communication cables, so that a plurality of first terminals and second terminals can be used to measure vibrations of a plurality of rotating machines and calculate vibration results. Since it should be configured, the installation cost is high, there was a problem that the volume of the system is large.

In addition, since wireless data transmission from the first terminal and the second terminal to the server is made in a network topology, there is a problem that it is difficult to diagnose and monitor a failure at a remote location due to an overload during wireless data transmission.

In order to solve the above problems, the present invention provides a fault diagnosis map of the GUI environment to prevent overload of wireless data transmission, and to grasp the fault diagnosis results of a plurality of machines at a glance, Provides USN-based rotating machine monitoring device and method using DPCM transmission method that can facilitate maintenance of each facility and prevent accidents caused by emergency trouble by enabling timely replacement and repair. The purpose is.

USN-based rotary machine monitoring apparatus using the DPCM transmission method according to the present invention to achieve the above object

A sensor unit 10 connected to one side of an industrial rotary machine and detecting vibration generated in a vertical or horizontal direction;

Connected to one side of the sensor unit converts the analog signal output from the sensor unit into a digital signal, collects the converted digital signal for a certain time, FFT conversion vibration data obtained through the wireless network to the Sink Node unit (Sink Node) DPCM type sensor node unit 20 to transmit,

A sink node unit 30 which receives the vibration data transmitted from the DPCM sensor node unit and transmits the vibration data to the remote server unit through RS-232 communication;

Monitor the vibration data of the DPCM sensor node of the remote node transmitted from the sink node, compare the existing vibration data with the existing experience data through the FFT fault diagnosis and analysis unit, and store it in a database or TCP / IP. It is achieved by being composed of a remote server unit 40 to transmit to a remote client PC through.

In addition, the USN-based rotary machine monitoring method using the PCM transmission method according to the present invention

Measuring the vibration of the industrial rotary machine by receiving the operation command of the DPCM-type sensor node in the sensor unit connected to one side of the industrial rotary machine (S100),

Converting an analog signal output from the sensor unit into a digital signal through a DPCM module of the DPCM type sensor node (S200);

The FCM converts the vibration data converted into the digital signal by the DPCM sensor node unit, generates the transmission packet by DPCM transmission method through the DPCM (Different Pulse Code Modulation) unit, and then transmits the vibration packet to the sink node unit. (S300),

Receiving the vibration data transmitted from the DPCM-type sensor node in the sink node unit and transmitting to the remote server unit by RF-232 communication (S400),

Under the control of the remote server unit, the vibration data from the remote DPCM sensor node transmitted from the sink node unit is monitored, and the existing vibration data are compared with the existing experience data through the FFT fault diagnosis and analysis unit, and stored in a database or This is achieved by the step (S500) of transmitting to a remote client PC via TCP / IP.

As described above, in the present invention, it is possible to quickly transmit data by preventing the overload of the wireless data transmission, thereby quickly diagnosing and analyzing the cause of the failure, and at a glance, it is possible to grasp the failure diagnosis results for a plurality of mechanical facilities. By providing a GUI map of the fault diagnosis of the GUI environment, it is possible to take timely measures such as replacement and repair of equipment depending on the result of the fault diagnosis.

In addition, the present invention has the advantage of stabilizing the overall equipment of the plant by strengthening maintenance by the root cause analysis of the failure of the mechanical equipment, thereby reducing the depreciation cost according to the extension of the machine life.

Recently, the concept of ubiquitous that combines physical space and electronic space has been widely used. Ubiquitous space refers to an environment in which not only people and computers but also people and objects, objects and objects are connected by wire and wireless, and communicate with each other and decide their own actions by inserting computers into all physical spaces.

As a specific example of such a ubiquitous environment, a sensor network based on various low-cost, low-power sensor nodes using a rapidly developing wireless communication technology and a semiconductor design technology may be mentioned.

The sensor network basically consists of a number of sensor nodes equipped with measurement and wireless communication functions.

These sensor nodes are able to form their own networks, and as a result, the measured information can be transferred to a backbone network such as the Internet through a self-formed network, which is widely applied to ecological monitoring, earthquake monitoring, and military use in a wide range of regions.

One of the best areas to take advantage of these sensor networks is in the field of fault diagnosis and analysis of mechanical equipment.

In the present invention, by applying the ubiquitous sensor network environment to the industrial rotary machine, it is characterized in that it is configured to analyze the failure causes for a plurality of industrial rotary machine, diagnose, and guide the result to the ubiquitous network.

In addition, in order to prevent overload of wireless data communication generated through the ubiquitous network, it is characterized by configuring a transmission packet through the DPCM transmission technique.

Hereinafter, preferred embodiments of the present invention will be described with reference to the accompanying drawings.

1 is a block diagram showing the components of the USN-based rotating machine monitoring apparatus using a DPCM transmission method according to the present invention, which is a sensor unit 10, DPCM type sensor node unit 20, sink node unit 30, the remote server unit 40.

First, the sensor unit 10 according to the present invention will be described.

The sensor unit 10 is connected to one side of the industrial rotating machine to detect the vibration generated in the vertical or horizontal direction, which consists of a temperature sensor, humidity sensor, illumination sensor, ultrasonic sensor, vibration sensor, acceleration sensor, It is installed in plural on the upper, lower, left and right sides of the part to be measured.

Industrial rotary machine described in the present invention refers to all devices that perform a rotary motion, as well as a motor, a cylinder shaft.

Next, the DPCM type sensor node unit 20 according to the present invention will be described.

The DPCM sensor node unit 20 according to the present invention is connected to one side of the sensor unit converts the analog signal output from the sensor unit into a digital signal, collects the converted digital signal for a predetermined time, vibration data obtained by FFT conversion Is transmitted to the Sink Node unit through a wireless communication network. As shown in FIG. 2, the DPCM module 21 and the message transmission packet unit 22 are configured.

The DPCM module 21 converts a signal output from the sensor unit into a digital signal, FFT converts the converted vibration data, and generates a transmission packet using a DPCM transmission technique through a differential pulse code modulation (DPCM) unit. As shown in FIG. 3, the digital signal converter 211 and the differential pulse code modulation (DPCM) unit 212 are configured.

The digital signal converter 211 converts an analog signal output from a sensor into a digital signal, which is an analog-digital converter having a 12-bit resolution.

For this reason, in the present invention, a signal output from the sensor can be used as an input and converted into a digital value of 0 to 4095.

The DPCM (Different Pulse Code Modulation) unit 212 converts only the PSD of the rotating device due to failure or abnormality instead of transmitting the entire PSD of the converted FFT to the remote sync node unit every prescribed time. It acts as a data abbreviation to transmit only specific data of contents.

That is, when A / D conversion and FFT conversion of vibration signal from acceleration sensor connected to DPCM, PSD (Power Spectrum Density) for each frequency band can be obtained, and facility failure diagnosis is based on these data. Is performed.

In the case of a rotating machine operating normally, these spectrum distributions are within a prescribed range, but when an abnormality occurs, the PSD value of each frequency band changes.

The remote sink node 30 receives the corresponding PSD data from the DPCM sensor node units mounted on each rotating machine, and the remote server unit connected to the sink node unit analyzes the received PSD data, Diagnose the failure of the rotating machine.

In this case, generally, the network topology is selected and configured, but the overhead for wireless data transmission becomes large, resulting in a decrease in transmission efficiency.

In order to solve this problem, in the present invention, instead of transmitting the entire PSD of the converted FFT at a prescribed time to the remote sink node unit, the changed content is specified only when the PSD of the rotating device is changed due to failure or abnormality. The DPCM transmits and generates DPCM transmission data through a differential pulse code modulation (DPCM) unit 212 to abbreviate and transmit only data.

In addition, the DPCM module 21 according to the present invention comprises a transmission buffer (TX_BUF) 213a for storing vibration data transmitted from the transmission node on one side of the transmission node 213 in order to transmit DPCM transmission data through the DPCM unit. Then, a receiving buffer (RX_BUF) 214b configured to store the result calculated using the received data and the previous data on one side of the receiving node 214 is configured.

With this configuration, when a change occurs in the PSD of the rotating machine, only the converted contents are stored in the transmission buffer (TX_BUF) 213 and the reception buffer (RX_BUF) 214 for transmission by the DPCM transmission technique.

The message transmission packet unit 22 transmits the vibration data obtained by the FFT conversion to the sink node unit 30 through the wireless communication network sequentially in a data message format.

The wireless communication network is composed of an RF communication network or a Zigbee communication network.

As shown in FIG. 5, the data area 225 of the TinyOS message 220 is modified to include an original message 225a and a DPCM message 225b.

The TinyOS message 220 includes an address unit 221 indicating whether the message is an RF message or a USART message, a type unit 222 indicating the type of the message, and a group ID. A group portion 223, a length portion 224 representing a length of the data region, a data region portion 225, which is a user-specified data region, and a CRC region portion 226 for error checking. It consists of

The original message 225a and the DPCM message 225b according to the present invention are modified and classified according to the structure of the data area 225 of the TinyOS message 220.

The original message 225a is transmitted once upon initial power-up, and serves to transmit all data generated as a result of the FFT. As shown in FIGS. 5 and 6, the source unit (Source) 225a-1, packet number section (Seqno) 225a-2, data area section 225a-3, ANP section 225a-4, and CPN section 225a-5.

That is, the source unit 225a-1 stores the ID of the transmission node, and the packet number unit 225a-2 stores the sequence number of the packet generation, and the data area unit 225a-3. A total of 13 pieces of data are stored in 2 bytes, and the ANP unit (All Packet Number) 225a-4 stores the number of all packets when the number of packets is multiple, and the current packet number (CPN) 225a-5. The number of the current packet among the total number of packets is stored.

The DPCM message 225b is used after the original message is transmitted, and serves to transmit data compressed by the DPCM technique, which is illustrated in FIGS. 5 and 7. As described above, the source section (225b-1), the packet number section (Seqno) 225b-2, the res section 225b-3, the DCC section 225b-4, and the data area section 225b-5 And an ANP unit 225b-6 and a CPN unit 225b-7.

The source unit 225b-1 stores the ID of the transmission node, and the packet number unit 225b-2 stores the sequence number of the packet generation, and the data area unit 225b-5. A total of 13 pieces of data are stored in 2 bytes, and the ANP unit (All Packet Number) 225b-6 stores the number of all packets when the number of packets is multiple, and the Current Packet Number (CPN) 225b-7 The number of the current packet among the total number of packets is stored.

 The Res unit 225b-3 is additionally used when the DCC bit is insufficient as the reserved bit. The DCC unit 225b-4 includes flag bits indicating whether the corresponding frequency is transmitted using a delta compression code.

In other words, if a value of '1' is recorded in the DCC bit, it indicates that the data value includes the Delta value for the corresponding frequency. If the value is '0', the data value does not include the Delta value of the frequency. do.

In the DCC code, flag values for a total of 40 frequencies are recorded. In addition, flag values for a total of 44 frequencies are recorded when a reserved bit is used.

Compressed data is stored in the data area, and since the compressed data has a size of 1 byte, a value between -128 and 127 is included.

Therefore, if the value exceeds -127 to 127, the Max value is recorded and transmitted. Since the value does not change suddenly, it is configured to gradually follow the target value.

As shown in FIG. 15, the operation of the message packet unit 22 according to the present invention checks whether a UART reception event has occurred, and receives a data from a remote DPCM sensor node if there is a UART reception event.

Subsequently, it is checked whether the received data is a DPCM message, and if it is a DPCM message, a TOS message based on a DPCM message is generated.

Next, the sink node unit 30 according to the present invention will be described.

The sink node unit 30 according to the present invention receives vibration data transmitted from the DPCM sensor node unit 20 and transmits the vibration data to the remote server unit through RS-232 communication.

It is composed of an RF communication network or a Zigbee communication network for wireless data communication with a DPCM sensor node unit of a remote location.

As shown in FIG. 16, the sink node unit checks whether there is an RF reception event from the DPCM sensor node unit at the remote location, receives data when there is an RF reception event, and transmits the data to the remote server unit through RS 232 communication. To send.

Here, the data refers to the diagnostic data of the industrial rotary machine measured by the DPCM sensor node of the remote site.

The sink node unit according to the present invention is connected to a remote server unit through RS-232 communication configured at one side, and transmits the vibration data transmitted from the DPCM sensor node unit to the remote server unit.

Next, the remote server unit 40 according to the present invention will be described.

The remote server unit according to the present invention monitors the vibration data transmitted from the DPCM sensor node of the remote node transmitted from the sink node unit, and compares the existing experience data with the current vibration data through the FFT fault diagnosis and analysis unit to analyze the data. It is a place to store or transmit to a remote client PC through TCP / IP. It is configured to notify the administrator's mobile phone when the device error of the remote server unit is connected due to CDMA connection on one side. And, the GUI program is configured to receive and monitor the vibration data transmitted from the remote DPCM sensor node unit.

In the remote server unit 40 according to the present invention, as shown in Fig. 4, the message processing unit 41, the web driving unit 42, and the FFT fault diagnosis and analysis unit 43 are constituted.

In addition, in order to receive the DPCM transmission data transmitted from the sink node, a reception buffer (RX_BUF) 44 is configured to store the result calculated using the received data and the previous data.

As shown in FIG. 16, the message processing unit 41 checks the message type transmitted from the sink node unit, and processes the original message to be immediately stored in the reception buffer (RX_BUF) and the database if the original message. Check to add only the changed data to the previous receive buffer (RX_BUF) and the contents stored in the database, and store the operation result in the receive buffer (RX_BUF) and the database.

The web driver 42 uses a web service of ASP.NET to request a failure diagnosis and analysis data of an industrial machine on a client PC connected by TCP / IP. It serves to provide analysis data.

The FFT fault diagnosis and analysis unit 43 compares the existing experience data with the current vibration data using the frequency spectrum data transmitted from the sink node unit, and analyzes the fault diagnosis and analysis data for the industrial machinery. It stores the data in the database.

As shown in FIG. 4, the FFT analyzer 43a converts a signal of a corresponding time domain for vibration data of a remote DPCM sensor node into a spectrum of a frequency domain;

After analysis through the FFT analyzer, the BP neural network of three layers (Input, Hidden, and Output) is formed for the region of interest of the fault, and for each fault diagnosis input data, It consists of a neural network unit 43b for diagnosing the failure through the output.

The FFT fault diagnosis and analysis unit 43 according to the present invention performs a fault diagnosis using the received vibration data in the remote server.

The failures of each motor corresponding to the rotating machine in the industrial facility may include various failures in addition to bearing failure, fan imbalance, structural looseness of the motor and shaft failure.

In order to diagnose such failures, the FFT failure diagnosis and analysis unit according to the present invention uses an FFT analysis that converts a signal in a time domain into a spectrum in a frequency domain.

8 to 11 are graphs showing the results of FFT conversion of vibration waveforms for bearing failures, fan imbalances, structural looseness of the motors, and shaft failures generated in rotating machines in industrial facilities.

In this way, the spectrum form is different for each failure, and the experience data is stored in advance in the database.

At this time, the FFT fault diagnosis and analysis unit according to the present invention is configured to perform fault diagnosis using a neural network unit for fault diagnosis.

In the present invention, the neural network unit is used for fault diagnosis because the region of interest of the fault can be easily input, and only the vibration data for the region of interest can be analyzed, so that the fault diagnosis can be easily analyzed.

FIG. 12 is a structural diagram showing the structure of a BP neural network applied to an FFT fault diagnosis and analysis unit according to the present invention, which illustrates a three-layer structure of an input layer, a hidden layer, and an output layer. Have

M in FIG. 12 is the number of neurons in the input layer, N is the number of neurons in the hidden layer, and L represents the number of neurons in the output layer.

Fault diagnosis is performed through the BP neural network applied to the FFT fault diagnosis and analysis unit according to the present invention.

For example, it is assumed that failures A, B, and C occurred due to long-term use of the equipment, and the FFT results of the measurement waveforms for the failures are as shown in FIG.

As shown in FIG. 13, each fault can be distinguished by a frequency magnitude between 0 and 20 Hz. In this case, the 0 to 20 Hz region is set as a region of interest.

The BP neural network of three layers (input layer, hidden layer, and output layer) is formed for the region of interest of the failure, and the program is set to output a corresponding output for each vibration data.

In other words, the input of BP network is inputted value of 0 ~ 20Hz, so there are 11 neurons in total and 3 output layer neurons (hidden layer neurons can be adjusted according to program setting).

The FFT result of the faults A, B, and C as shown in FIG. 13 is input to the generated network, and the program is set to have a desired output.

That is, when the frequency magnitude value between 0 and 20 Hz for fault A is used as an input, the output value is set to [1 0 0], and in case of fault B and fault C, the output is [0 1 0]. And [0 0 1] appear.

When the output is completed, a new region of interest of the FFT value for the newly measured input signal is newly input to the current network.

Assume that the output for the new input is [0.7 0.3 0.2]. Then, in the present invention, the FFT failure diagnosis and analysis unit analyzes that the probability of failure A is 70%, the probability of failure B 30%, and the probability of failure C 20%.

Hereinafter, the USN-based rotary machine monitoring method using the DPCM transmission method according to the present invention will be described in detail.

14 is a flowchart illustrating a USN-based rotating machine monitoring method using a DPCM transmission technique according to the present invention.

First, by receiving the operation command of the DPCM-type sensor node in the sensor unit connected to one side of the industrial rotary machine to measure the vibration of the industrial rotary machine (S100).

The industrial rotary machine is set as a motor, the vibration sensor is installed on the upper, lower, left, right surface of the motor.

Next, an analog signal output from the sensor unit is converted into a digital signal through the DPCM module of the DPCM type sensor node (200).

It uses a 12-bit resolution analog-to-digital converter to convert signals from the sensor to digital values from 0 to 4095.

Next, the FCM converts the vibration data converted into the digital signal in the DPCM sensor node unit, generates the transmission packet by the DPCM transmission technique through the DPCM (Different Pulse Code Modulation) unit, and then sends it to the sink node unit. To transmit (S300).

The DPCM (Different Pulse Code Modulation) unit transmits the entire PSD of the transformed FFT every predetermined time to the remote sync node unit instead of transmitting specific PSD data to the rotating device due to failure or abnormality. Abbreviate data transfers only.

First, a process of FFT transforming vibration data converted into a digital signal in a DPCM sensor node unit and a process of first transmitting FFT transformed vibration data to a sink node unit using a DPCM transmission method will be described.

As an example, it is assumed that N = 128, which is the number of samples, and that the sampling frequency is 100 Hz (T = 1/100).

At this time, the variable frequency and the frequency calculation can be expressed by the following equations (1) and (2).

That is, since the sampling frequency is 100 Hz, the analysis can be performed up to 50 Hz, and the threshold for comparing the variation amount is set to 10.

At this time, the DPCM sensor node according to the present invention is operated as follows for FFT conversion.

As shown in FIG. 15, the DPCM sensor node according to the present invention measures N sample data and then performs an FFT operation (using a real signal FFT) (S311 to S312).

In this case, magnitude values for a total of 64 frequency components are generated as a result of the calculation generated by the FFT.

Since both the first transmit buffer TX_BUF value and the receive buffer RX_BUF value are 0, the FFT calculation result is configured as the data area of the original message (S313).

Subsequently, an original message is generated through the message transmission packet unit, and all are transmitted to the sink node unit through the Zigbee module which is a wireless communication network (S321 and S323).

At this time, data transmission from the DPCM sensor node unit to the sink node unit is performed as follows.

The total data generated as a result of the calculation is 64, and since the original message can contain a total of 13 data, the total message consists of 5 messages.

That is, in the DPCM sensor node unit, a message including a part (Source, Seqno, Data, APN, CPN) and type information corresponding to the data area of the TinyOS message is composed and transmitted to the sink node unit.

After generating the TinyOS message, the sink node records the received data and type in the corresponding area, and generates the final message by recording the corresponding information in the remaining address part, group part, length part, and CRC area part.

In this case, since the total number of transmitted messages is 5, the APN of the message is 5, and the CPN is recorded and transmitted from 1 to 5 according to the order of the messages.

When the transmission is completed, the transmitted vibration data is recorded in the receiving buffer portion of the DPCM sensor node portion.

As described above, when the FCM operation (using a real signal FFT) is performed for the first time in the DPCM sensor node, the DPCM packet can be expressed as shown in Tables 1 to 3 below.

[DCC area of DPCM packet 1] N One 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 One 0 0 0 One 0 One 0 0 0 0 0 0 One 0 0 0 0 One N 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 0 0 0 0 0 One 0 0 0 One 0 One 0 0 0 0 0 0 One

N 39 40 41 42 43 44 0 0 0 0 0 0

[Data Area of DPCM Packet 1] One 2 3 4 5 6 7 30 -20 40 20 40 60 -57

[APN / CPN in DPCM Packet 1] APN CPN 2 One

Subsequently, FFT operation is performed on the next N measured sample data following the DPCM sensor node part, and the difference data is calculated by using the vibration data generated as a result of the FFT operation and previous data transmitted in the previously received receiving buffer part. It is checked whether the value of the frequency band has changed (S313 to S314).

Subsequently, the DCC value of the DPCM message is calculated using the change data, and the data area of the DPCM message is constructed using the DCC value (S315 to S316).

Subsequently, the message transmission packet unit 22 transmits the data area of the DPCM message (S317).

Subsequently, the message transmission packet unit 22 checks whether a UART reception event has occurred, and receives a data if there is a UART reception event (S318 to S319).

Subsequently, it is checked whether the received data is a DPCM message (S320). If the DPCM message is generated, a TOS message based on a DPCM message is generated (S322). In operation S323, the data is transmitted to the sink node.

Subsequently, it is checked whether the transmission is completed to the message transmission packet part, and when the transmission is completed, the transmission buffer is corrected (S324 to S325).

For example, assuming that the value of the frequency band of the difference data is changed (the change data means a value larger than the threshold value 10 assumed above), using the converted data as shown in Table 4 below. The DCC shown in Tables 5 to 7 can be prepared, and DPCM result data can be recorded in the data area based on the DCC.

And, generating the transmission packet by the DPCM transmission technique through the DPCM (Different Pulse Code Modulation) unit,

The ID of the transmitting node is stored in the Source 225b-1 of the DPCM Message 225b, and the packet number section 225b-2 stores the sequence number of the packet generation. If the DCS bit is insufficient as the reserved bit in the Res unit 225b-3, the bit is additionally used, and flag bits indicating whether the corresponding frequency is transmitted as the Delta compression code are displayed in the DCC unit 225b-4, and the data is displayed. Stores 13 pieces of data each by 2 bytes in the area unit 225b-5, and stores the number of all packets when there are multiple packets in the ANP unit (All Packet Number) 225b-6, and CPN (Current Packet) Number) 225b-7 stores the number of the current packet among the total number of packets to generate a transmission packet.

Magnitude for Changed Frequency N One 5 7 14 19 25 34 47 △ 30 -20 40 20 40 60 -57 72

[DCC area of DPCM packet 2] N One 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 0 0 One 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 N 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

N 39 40 0 0

[DCC area of DPCM packet 2] One 72

[APN / CPN in DPCM Packet 2] APN CPN 2 2

That is, as shown in Table 4, since the number of changed frequencies is eight, one message may include a data area. However, when transmitting a message, the frequency information may be transmitted by using a DCC value without separately transmitting a frequency value. Since we are sending it, we create two messages and send them.

As shown in Table 1, looking at the DCC region of DCPM packet 1, the DCC value has a total of 44 bits up to the reserved bit. Therefore, '0' or '1' is recorded in the value of the corresponding bit to indicate whether the size information for the 44th frequency is included in the data from the size value for the first frequency.

As a result, as shown in Table 2, the number of messages is two, and the first message, DCPM packet 1, contains values corresponding to 1, 5, 7, 14, 19, 25, and 34.

In addition, the second message DCPM packet 2 includes a value corresponding to the 47th frequency.

That is, as shown in Table 5, when looking at the DCC region of DCPM packet 2, it can be seen that only the bit corresponding to the changed frequency is set to '1', and only the change amount value corresponding to the frequency set to '1' is the data region. You can see it included in.

The seqno part is increased by 1, and the total number of messages is 2, so the APN is 2, and the CPN is 1 when the first message is 2, and 2 is the second message.

Through this process, the difference between the magnitude value of the frequency component generated as a result of the FFT operation and the data of the previously transmitted receiving buffer part is calculated, and the calculated vibration data is compressed by the DPCM transmission method and transmitted to the sink node. .

Next, the sink node receives the vibration data transmitted from the DPCM-type sensor node unit and transmits to the remote server unit in the RS-232 communication (S400).

This checks whether or not the UART reception event has occurred, and receives the vibration data from the DPCM sensor node unit of the remote if there is a UART reception event (S410 ~ S411).

Then, it checks whether the received vibration data is a DPCM message, and if the DPCM message, generates a TCM message based on the DPCM message, and if not the DPCM message generates a message based on the original message and transmits the vibration data to the remote server unit. (S412).

Next, under the control of the remote server unit, the vibration data from the remote DPCM sensor node transmitted from the sink node unit is monitored, and the existing experience data and the current vibration data are compared and analyzed through the FFT fault diagnosis and analysis unit. Store or transmit to the remote client PC via TCP / IP (S500).

This checks whether the serial reception event is generated from the sink node via RS-232 communication as shown in FIG.

Subsequently, when the serial reception event occurs, the vibration data transmitted from the sink node is received, and it is checked whether the vibration data is a DPCM message (S511).

Subsequently, in the case of the DPCM message, the data is restored using the previous received buffer value and the currently received vibration data (S513), and the received buffer value is corrected (S514).

The process of comparing and analyzing existing experience data and current vibration data through the FFT fault diagnosis and analysis unit is performed as follows.

First, the vibration data of the remote DPCM sensor node is transformed into the spectrum of the frequency domain through the FFT analyzer of the FFT fault diagnosis and analysis unit.

In other words, vibration data on bearing failure, fan imbalance, structural looseness of the motor, shaft failure, etc. generated by the motor corresponding to the rotating machine in the industrial equipment is converted into the spectrum of the frequency domain through the FFT analyzer. Convert

The signal of the corresponding time domain is set to a frequency domain generated between 0 and 20 Hz.

Subsequently, after analysis through an FFT analyzer, a BP neural network of three layers (input, hidden and output) is formed for the region of interest of the failure, and corresponding vibration data is generated. Diagnose the fault through the output.

For example, it is assumed that failures A, B, and C occurred due to long-term use of the equipment, and the FFT results of the measurement waveforms for the failures are as shown in FIG.

As shown in FIG. 12, each fault can be distinguished by a frequency magnitude between 0 and 20 Hz. In this case, the 0 to 20 Hz region is set as a region of interest.

The BP neural network of three layers (input layer, hidden layer, and output layer) is formed for the region of interest of the failure, and the program is set to output a corresponding output for each vibration data.

In other words, the input of BP network is inputted value of 0 ~ 20Hz, so there are 11 neurons in total and 3 output layer neurons (hidden layer neurons can be adjusted according to program setting).

The FP result of the failures A, B, and C as shown in FIG. 13 is input to the generated BP network, and the program is set so that a desired output is obtained.

That is, when the frequency magnitude value between 0 and 20 Hz for fault A is used as an input, the output value is set to [1 0 0], and in case of fault B and fault C, the output is [0 1 0]. And [0 0 1] appear.

When the output is completed, a new region of interest of the FFT value for the newly measured input signal is newly input to the current network.

Assume that the output for the new input is [0.7 0.3 0.2]. Then, in the present invention, the FFT failure diagnosis and analysis unit analyzes that the probability of failure A is 70%, the probability of failure B 30%, and the probability of failure C 20%.

1 is a block diagram showing the components of a USN-based rotating machine monitoring apparatus using a DPCM transmission technique according to the present invention,

2 is a block diagram showing the components of the DPCM sensor node unit 20 according to the present invention;

3 is a block diagram showing the components of the DPCM module 21 according to the present invention;

4 is a block diagram showing components of a remote server unit according to the present invention;

5 is a structural diagram showing the components of the message transmission packet unit according to the present invention;

FIG. 6 illustrates an original message 22a structure that is transmitted once upon initial power-up among the components of the message transmission packet unit 22 according to the present invention, and transmits all data generated as a result of the FFT. In one embodiment,

7 is a DPCM message 22b which is used after the original message is transmitted among the components of the message transmission packet unit 22 according to the present invention and transmits compressed data by the DPCM technique. An embodiment showing a structure)

FIG. 8 is a graph showing empirical data storing winding failures of motors in an industrial rotating machine through FFT analysis.

FIG. 9 is a graph showing empirical data in which structural looseness of a motor in an industrial rotating machine is stored through FFT analysis.

FIG. 10 is a graph showing empirical data in which unbalance of a motor shaft of an industrial rotating machine is stored through FFT analysis; FIG.

FIG. 11 is a graph showing empirical data storing unbalanced rotating elements in an industrial rotating machine through FFT analysis. FIG.

12 is a structural diagram showing a structure of a BP neural network applied to an FFT fault diagnosis and analysis unit according to the present invention;

FIG. 13 is a graph showing output data obtained by inputting FFT results for failures A, B, and C into a BP neural network applied to an FFT failure diagnosis and analysis unit according to the present invention;

14 is a flowchart illustrating a USN-based rotating machine monitoring method using a DPCM transmission technique according to the present invention;

15 is a flow chart showing an operation process in the DPCM type sensor node unit 20 according to the present invention,

16 is a flowchart illustrating an operation process in a sink node unit and a remote server unit according to the present invention.

※ Brief description of reference numerals ※

10: sensor unit 20: DPCM type sensor node

21: DPCM module 22: message transmission packet unit

30: sink node 40: remote server

41: message processing section 42: web driver

43: FFT fault diagnosis and analysis unit 211: digital signal conversion unit

212: DPCM part 225a: original message

225b: DPCM message

Claims (7)

A sensor unit 10 connected to one side of an industrial rotary machine and detecting vibration generated in a vertical or horizontal direction; Connected to one side of the sensor unit converts the analog signal output from the sensor unit into a digital signal, collects the converted digital signal for a certain time, FFT conversion vibration data obtained through the wireless network to the Sink Node unit (Sink Node) A DPCM type sensor node unit 20 including a DPCM module 21 and a message transmission packet unit 22 for transmitting; A sink node unit 30 which receives the vibration data transmitted from the DPCM sensor node unit and transmits the vibration data to the remote server unit through RS-232 communication; Monitor the vibration data of the DPCM sensor node of the remote node transmitted from the sink node, compare the existing vibration data with the existing experience data through the FFT fault diagnosis and analysis unit, and store it in a database or TCP / IP. In the configuration of the remote server unit 40 for transmitting to a remote client PC via, The DPCM module 21 stores only the converted contents in the transmit buffer (TX_BUF) 213 and the receive buffer (RX_BUF) 214 when a change occurs in the PSD of the rotating machine. USN-based rotating machine monitoring apparatus using the DPCM transmission method, characterized in that for transmitting to the DPCM transmission method through (212). delete delete The method of claim 1, wherein the message transmission packet unit 22 An original message 22a transmitted once upon initial power-up and transmitting all data generated as a result of the FFT; USN-based rotary machine using DPCM transmission technique, which consists of DPCM message 22b which is used after original message is transmitted and transmits compressed data by DPCM technique. Monitoring device. The remote server unit 40 of claim 1, wherein the remote server unit 40 converts a signal in a corresponding time domain with respect to FFT spectrum data of a remote DPCM sensor node into a spectrum in a frequency domain, and through the FFT analyzer. After analysis, BP neural network of three layers (Input, Hidden, and Output) is formed for the region of interest of the fault, and the fault is output through the corresponding output for each fault diagnosis input data. USN-based rotary machine monitoring apparatus using a DPCM transmission method, characterized in that the configuration comprises a FFT fault diagnosis and analysis unit 43 consisting of a neural network unit (43b) to diagnose. Measuring the vibration of the industrial rotary machine by receiving the operation command of the DPCM-type sensor node in the sensor unit connected to one side of the industrial rotary machine (S100), Converting an analog signal output from the sensor unit into a digital signal through a DPCM module of the DPCM type sensor node (S200); The FCM converts the vibration data converted into the digital signal by the DPCM sensor node unit, generates the transmission packet by DPCM transmission method through the DPCM (Different Pulse Code Modulation) unit, and then transmits the vibration packet to the sink node unit. (S300), Receiving the vibration data transmitted from the DPCM-type sensor node in the sink node unit and transmitting to the remote server unit by RF-232 communication (S400), Under the control of the remote server unit, the vibration data from the remote DPCM sensor node transmitted from the sink node unit is monitored, and the existing vibration data are compared with the existing experience data through the FFT fault diagnosis and analysis unit, and stored in a database or In the step (S500) of transmitting to a remote client PC via TCP / IP, The generation of the transmission packet by the DPCM transmission technique through the DPCM (Different Pulse Code Modulation) unit The ID of the transmitting node is stored in the Source 225b-1 of the DPCM Message 225b, and the packet number section 225b-2 stores the sequence number of the packet generation. If the DCS bit is insufficient as the reserved bit in the Res unit 225b-3, the bit is additionally used, and flag bits indicating whether the corresponding frequency is transmitted as the Delta compression code are displayed in the DCC unit 225b-4, and the data is displayed. Stores 13 pieces of data each by 2 bytes in the area unit 225b-5, and stores the number of all packets when there are multiple packets in the ANP unit (All Packet Number) 225b-6, and CPN (Current Packet) USN-based rotating machine monitoring method using the DPCM transmission method, characterized in that for generating the transmission packet by storing the number of the current packet among the total number of packets in the number (225b-7). delete

Images