CN104811373A - Multifunctional real-time synchronous acquisition internet of things gateway - Google Patents

Abstract

The invention relates to a multifunctional real-time synchronous acquisition internet of things gateway. In allusion to an actual requirement for signal synchronous acquisition in the field of mechanical equipment fault diagnosis, unit fault monitoring and diagnosing effects are prevented from being seriously affected by phase deviation. The internet of things gateway also has prominent functions of rotating speed prediction, advanced filtering and data synchronous correction and processing on the basis of having wide access abilities and typical management and protocol conversion functions. The internet of things gateway can acquire various physical quantities and electric signals outputted by sensors in real time, the sensors comprise a temperature sensor, a pressure sensor, a flow sensor, an acceleration sensor, a speed sensor, an eddy current sensor, a rotating speed sensor and the like. Meanwhile, the internet of things gateway can also complete data communication with other systems. The internet of things gateway can carry out precise processing and analysis and calculate various characteristic values on the basis of completing data acquisition and synchronous processing, wherein the characteristic values comprise a vibration velocity virtual value, a vibration acceleration true peak value, a vibration displacement true peak value, various frequency doubling values of vibration signals, a GIE value and the like.

Description

A kind of multi-functional real-time synchronization gathers things-internet gateway

Technical field

The present invention relates to technology of Internet of things field, particularly relate to a kind of multi-functional real-time synchronization and gather things-internet gateway.

Background technology

Under the operational environment of gateway of internet of things is generally all in unattended operation condition, complete rear gateway is tentatively set starts to carry out work, following it will become the tie connecting sensing network and conventional communication networks.As gateway device, things-internet gateway can realize sensing network and communication network, and the connection between dissimilar sensing network, both can realize wide area interconnected, also can realize Interconnected.

Transducer is the basic components of Internet of Things, typical transducer involving vibrations transducer, temperature sensor, pressure sensor, flow sensor etc.The signal demand that different transducers gathers, through data acquisition processing device, changes digital signal into by analog electric signal, thus carries out transmitting, store and analyzing.

In mechanical fault diagnosis field, things-internet gateway is connected with various transducer, unit real-time running state is obtained from transducer, equipment fault requires higher to signal synchronous collection, once there is deviation or unit malfunction monitoring diagnosis effect will be had a strong impact on unit actual rotation phase place is inconsistent in unlike signal phase place, the generation of the serious accidents such as setting causes seriously, Fire explosion.

From the actual demand of mechanical Fault Monitoring of HV diagnosis, this things-internet gateway, also can the synchronous process of settling signal except possessing widely except access capability and typical management, protocol conversion function.This things-internet gateway can the various physical quantity that exports of Real-time Collection transducer and the signal of telecommunication, comprises the transducers such as temperature, pressure, flow, acceleration, speed, current vortex, rotating speed, simultaneously can also complete data communication with other system; At data acquisition with on the synchronous basis processed, also can carry out fine processing analysis, calculate various characteristic value, involving vibrations speed effective value, vibration acceleration real peak value, vibration displacement real peak peak value, each frequency multiplication value of vibration signal, GIE value etc.

This gateway, using the key node as mechanical fault diagnosis data transmission and processing, plays a significant role in fault diagnosis field.

Summary of the invention

A kind of multi-functional real-time synchronization gathers things-internet gateway, it is characterized in that:

This gateway comprises with lower unit: real time data acquisition unit, generating date unit, real time data synchronization processing unit, CPU processing unit, Data Post unit, data transmission unit, data real-time reception unit;

Above-mentioned each unit connection relation is: real time data acquisition unit is connected with generating date unit, and the Various types of data after gathering is transferred to generating date unit; Generating date unit is connected with real time data acquisition unit, real time data synchronization processing unit, and the data of reception are completed filtering by generating date unit, intercepting complete cycle is transferred to real time data synchronization processing unit after processing; Real time data synchronization processing unit is connected with real time data synchronization processing unit, CPU processing unit, and real time data synchronization processing unit is transferred to CPU processing unit after the data of reception are carried out synchronous correction process; CPU processing unit is connected with real time data synchronization processing unit, Data Post unit, data real-time reception unit, data transmission unit respectively, Data Post unit has been used for data feature values and has extracted, data real-time reception unit can receive the data that other equipment gathers, and data transmission unit is used for sending the data to Internet of Things management platform.

Further, generating date unit carries out the dynamic linear low-pass filtering of 20 times of power frequencies of real-time continuous to data waveform;

Continuous print 20 times of power frequency filtering need the corresponding rotating speed predicting waveform in advance, choose the filter of corresponding flex point in advance, and will accomplish the seamless switching of filter according to this rotating speed, and adopt FPGA to do real-time flexible switching and filtering, method is as follows:

1) according to the prediction rotating speed that history rotating speed obtains next time through recursive operation, prediction rotating speed computing formula is as follows:

Wherein X' _nthe prediction rotating speed of n-th time, K _n-1the correction coefficient of (n-1)th time, X _n-1the actual speed of (n-1)th time, (n-1)th Ratio for error modification; β is unit driving characteristic parameter, and its computational methods are as follows:

$\mathrm{\β}=\frac{\underset{i=1}{\overset{\mathrm{num}\left(Y\right)}{\mathrm{\Σ}}}{Y_i}^{\′\′}}{\mathrm{num}\left(Y\right)}$

Wherein Y is unit startup procedure rotary speed data, and num (Y) is generating unit speed number of data points;

The correction coefficient K of n-th time _ncomputing formula is as follows:

$K_n=\frac{\sqrt{\underset{i=1}{\overset{5}{\mathrm{\Σ}}}{\left(\frac{X_{n-i}}{X_{n-i-1}}\right)}^2}}{5}$

The Ratio for error modification of n-th time computing formula is as follows:

2) using the filter of corresponding flex point according to prediction rotating speed, in order to realize the seamless connection of different filter, adopting multiple filter filtering simultaneously, according to the Output rusults of the direct gating respective filter of prediction rotating speed;

Further, it is characterized in that the data synchronization processing mode of real time data synchronization processing unit;

(1) adopt analog circuit phase-correction filter to carry out preliminary synchronisation to data waveform, analog circuit phase-correction filter has the characteristic with the complementation of simulated modulation filter, group delay can be corrected to close to linear curve;

(2) data after preliminary synchronisation process will in the following way, processes:

1), after acceleration passage carries out continuous 10KHz low pass linear filtering, 25.6KHz is resampled to;

2), after displacement passage carries out continuous 2KHz low pass linear filtering, 5.12KHz is resampled to;

3) after speed channels carries out continuous 1KHz low pass linear filtering, be resampled to 2.56KHz, then enter integrating circuit and complete integration, and then carry out phasing by the phase correction module of a digital circuit, offset the phase deviation of different frequency;

4), after stress, strained channel carry out continuous 1KHz low pass linear filtering, 2.56KHz is resampled to;

5), after traffic channel carries out continuous 500Hz low pass linear filtering, 1KHz is resampled to;

6), after pressure channel carries out continuous 500Hz low pass linear filtering, 1KHz is resampled to.

Accompanying drawing explanation

Fig. 1: each unit connection relation figure of the present invention;

Fig. 2: embodiment of the present invention real time data synchronization processing unit cut-away view;

Fig. 3: embodiment of the present invention generating unit speed simulation result figure;

Fig. 4: embodiment of the present invention generating date unit multistage filter structure figure;

Embodiment

As shown in Figure 1, a kind of multi-functional real-time synchronization gathers things-internet gateway and comprises following 7 unit: real time data acquisition unit, generating date unit, real time data synchronization processing unit, CPU processing unit, Data Post unit, data transmission unit, data real-time reception unit.

Real time data acquisition unit is connected with generating date unit, and the Various types of data after gathering is transferred to generating date unit; Generating date unit is connected with real time data acquisition unit, real time data synchronization processing unit, and the data of reception are completed filtering by generating date unit, intercepting complete cycle is transferred to real time data synchronization processing unit after processing; Real time data synchronization processing unit is connected with real time data synchronization processing unit, CPU processing unit, and real time data synchronization processing unit is transferred to CPU processing unit after the data of reception are carried out synchronous correction process; CPU processing unit is connected with real time data synchronization processing unit, Data Post unit, data real-time reception unit, data transmission unit respectively, Data Post unit has been used for data feature values and has extracted, data real-time reception unit can receive the data that other equipment gathers, and data transmission unit is used for sending the data to Internet of Things management platform.

Each unit concrete function is as described below:

real time data acquisition unit: the unlike signal that the various transducer of Real-time Collection exports, comprises the transducers such as temperature, pressure, flow, acceleration, speed, current vortex, rotating speed;

generating date unit: for processing the various data collected in real time, real time process comprises the dynamic linear low-pass filtering of 20 times of power frequencies of in real time data waveform being carried out to real-time continuous; Key signal is used data waveform to be carried out to the waveform interception of real-time continuous, the vibration in several cycles of accurate reproduction unit, impact, pressure, flow, temperature and displacement information;

real time data synchronization processing unit: the data that during for logarithm factually, processing unit processes is crossed carry out synchronous correction;

cPU processing unit: for by Embedded open operating system controlling application program deal with data, and make certain judgement according to system convention; If data exist abnormal, import problem data waveform into data transmission unit continuously and give Internet of Things management platform; If data are not abnormal, then the characteristic value that a transmission processing is complete, thus reduce network load;

data Post unit: for carrying out fine processing analysis to image data, be calculated as follows characteristic value: effective value, real peak value, real peak peak value, effective peak, effectively peak-to-peak value, each frequency multiplication value, GIE value, be still transferred to CPU processing unit after characteristic value process completes;

data transmission unit: send the data to Internet of Things management platform by WIFI, fiber optic Ethernet or Ethernet;

data real-time reception unit: for receiving the data that other equipment gathers, can receive and comprise CAN, the data of Zigbee bus and Modbus bus

The continuous print that generating date unit completes 20 times of power frequency filtering need the corresponding rotating speed predicting waveform in advance, the filter of corresponding flex point is chosen in advance according to this rotating speed, and accomplish the seamless switching of filter, adopt FPGA to do real-time flexible switching and filtering, method is as follows:

1) the prediction rotating speed obtaining next time according to history rotating speed through recursive operation by FPGA, prediction rotating speed computing formula is as follows:

Wherein X' _nthe prediction rotating speed of n-th time, K _n-1the correction coefficient of (n-1)th time, X _n-1the actual speed of (n-1)th time, (n-1)th Ratio for error modification; β is unit driving characteristic parameter, and its computational methods are as follows:

$\mathrm{\β}=\frac{\underset{i=1}{\overset{\mathrm{num}\left(Y\right)}{\mathrm{\Σ}}}{Y_i}^{\′\′}}{\mathrm{num}\left(Y\right)}$

Wherein Y is unit startup procedure rotary speed data, and num (Y) is generating unit speed number of data points;

The correction coefficient K of n-th time _ncomputing formula is as follows:

$K_n=\frac{\sqrt{\underset{i=1}{\overset{5}{\mathrm{\Σ}}}{\left(\frac{X_{n-i}}{X_{n-i-1}}\right)}^2}}{5}$

The Ratio for error modification of n-th time computing formula is as follows:

Select a unit in embodiment, the part-time point data of its actual speed is as shown in table 1.

Table 1 unit actual speed data

According to above-mentioned algorithm, emulation obtains unit simulation rotary speed data, and the data of part-time point are as shown in table 2.

Table 2 unit simulation rotary speed data

Sequence number	1	2	3	4	5	6	7	8	9	10
											Emulation rotating speed	0	-0.07	1.29	6.20	15.07	24.60	34.07	46.27	65.69	84.84
Sequence number	131	132	133	134	135	136	137	138	139	140
											Emulation rotating speed	1074.43	945.62	792.82	646.57	511.68	396.76	298.63	218.42	151.88	99.16

Compare unit actual speed and simulation rotary speed data, draw unit speed diagram as shown in Figure 2, solid line represents actual speed, and what dotted lines represented is simulation rotating speed.

2) using the filter of corresponding flex point according to prediction rotating speed, in order to realize the seamless connection of different filter, adopting multiple filter filtering simultaneously, according to the Output rusults of the direct gating respective filter of prediction rotating speed; As shown in Figure 3, in Fig. 3, left side is 5 groups of filters to filter construction, and flex point is 5 groups of parameters near prediction flex point, and wherein L is stepping coefficient, can mate according to different machine unit characteristic.

The sample rate adopted due to unlike signal kind is different, and generating date unit has carried out various filters operation to each circuit-switched data waveform, every road waveform can be caused all to produce phase shifts, can affect greatly the data processing in later stage and judgement after waveform loses phase place; Real time data synchronization processing unit adopts digital filter to correct these skews, to reach the effect of phase place maintenance strict conformance after all passage process.

Real time data synchronization processing unit is made up of following two parts:

analog circuit phase correction module

digital filter phase correction module

Real time data synchronization processing unit internal structure as shown in Figure 4.Because simulated modulation part of the present invention adopts Butterworth filter to do anti-aliasing filtering, this filter has the advantages that structure is simple, gain is large, but brings the shortcoming of nonlinear phase simultaneously.Nonlinear phase can cause the phase shift of different frequency inconsistent, and phase information there will be error, and when some fault occurring and occurring, these errors may cause diagnostic result inaccurate.Real time data synchronization processing unit, first according to the transfer function of front end modulate circuit, generate an analog circuit phase-correction filter for this passage, this analog circuit phase-correction filter has the characteristic with the complementation of simulated modulation filter, group delay can be corrected to close to linear curve.

Data, according to predefined configuration, are divided into following processing mode by the data waveform after above-mentioned primary sync process.

after carrying out continuous 10KHz low pass linear filtering, be resampled to 25.6KHz, this channel data carries out reprocessing as Acceleration pulse;

after carrying out continuous 2KHz low pass linear filtering, be resampled to 5.12KHz, this channel data carries out reprocessing as displacement waveform;

after carrying out continuous 1KHz low pass linear filtering, be resampled to 2.56KHz, and then enter integrating circuit, because integrating circuit is nonlinear, therefore integrating circuit back adds the phase correction module of a digital circuit more again, offset the phase deviation of different frequency, this channel data carries out reprocessing as velocity wave form.

ess-strain passage is resampled to 2.56KHz after carrying out continuous 1KHz low pass linear filtering;

traffic channel is resampled to 1KHz after carrying out continuous 500Hz low pass linear filtering;

pressure channel is resampled to 1KHz after carrying out continuous 500Hz low pass linear filtering.

A kind of multi-functional real-time synchronization that the present invention relates to gathers third party's data-interface that things-internet gateway can configure at most No. 64 vibrating sensors and magnanimity, therefore the operational capability of processing unit is most crucial with optimization in real time, and gateway normally runs to adopt multiple thinking to ensure in the present invention.

1) the tms320dm8168 chip adopting TI up-to-date, its kernel is ARM A9 processor, adopts Davinci structure, has the DSP of an ARM and 1GHz of a 1GHz dominant frequency.DSP primary responsibility completes calculation function, and ARM uses as scheduling unit, and at DSP just when computing, next group data is carried out preliminary treatment by ARM, under whole system being operated in be similar to the pattern of streamline, can increase system operations throughput greatly.

2) three bufferings of ARM+DDR, FPGA+DDR and ARM+NAND are adopted.Wherein ARM+DDR has the buffering area of 1GB, and FPGA+DDR has the buffering area of 4GB, and ARM+NAND has 32GB buffering area.When whole system operational capability can complete real-time process needs, complete data buffering by ARM+DDR.When fast detecting report is abnormal, passage is reported to the police or expert diagnostic system automatically gets involved time, system enters thin analytical model, at this moment system operations amount is increased sharply, processor cannot complete the computing of magnanimity, at this time except ARM+DDR buffer, FPGA+DDR buffer is also got involved in time and is avoided overflowing, and when the free space of FPGA+DDR buffer reaches warning critical line, ARM+NAND buffer is got involved and adapter data.Until store data gradually digested fall, ARM+NAND and FGPA+DDR exits intervention gradually, recover standby mode.

3) in order to avoid after three buffers get involved simultaneously, data accumulation too much causes seeing up-to-date data, the design have employed the pattern of 3+1 to ensure that up-to-date data waveform is uploaded, namely 3 groups of current datas are often processed, get one group of up-to-date data carry out computing and upload, so both ensure that the segmentation of data is analysed, reach again the demand of viewing in real time.

Claims (3)

1. multi-functional real-time synchronization gathers a things-internet gateway, it is characterized in that:

This gateway comprises with lower unit: real time data acquisition unit, generating date unit, real time data synchronization processing unit, CPU processing unit, Data Post unit, data transmission unit, data real-time reception unit;

Above-mentioned each unit connection relation is: real time data acquisition unit is connected with generating date unit, and the Various types of data after gathering is transferred to generating date unit; Generating date unit is connected with real time data acquisition unit, real time data synchronization processing unit, and the data of reception are completed filtering by generating date unit, intercepting complete cycle is transferred to real time data synchronization processing unit after processing; Real time data synchronization processing unit is connected with real time data synchronization processing unit, CPU processing unit, and real time data synchronization processing unit is transferred to CPU processing unit after the data of reception are carried out synchronous correction process; CPU processing unit is connected with real time data synchronization processing unit, Data Post unit, data real-time reception unit, data transmission unit respectively, Data Post unit has been used for data feature values and has extracted, data real-time reception unit can receive the data that other equipment gathers, and data transmission unit is used for sending the data to Internet of Things management platform.

2. gather things-internet gateway according to the multi-functional real-time synchronization of the one described in claim 1, it is characterized in that: generating date unit carries out the dynamic linear low-pass filtering of 20 times of power frequencies of real-time continuous to data waveform;

1) according to the prediction rotating speed that history rotating speed obtains next time through recursive operation, prediction rotating speed computing formula is as follows:

Wherein X' _nthe prediction rotating speed of n-th time, K _n-1the correction coefficient of (n-1)th time, X _n-1the actual speed of (n-1)th time, (n-1)th Ratio for error modification; β is unit driving characteristic parameter, and its computational methods are as follows:

Wherein Y is unit startup procedure rotary speed data, and num (Y) is generating unit speed number of data points;

The correction coefficient K of n-th time _ncomputing formula is as follows:

The Ratio for error modification of n-th time computing formula is as follows:

2) using the filter of corresponding flex point according to prediction rotating speed, in order to realize the seamless connection of different filter, adopting multiple filter filtering simultaneously, according to the Output rusults of the direct gating respective filter of prediction rotating speed.

3. gather things-internet gateway according to the multi-functional real-time synchronization of the one described in claim 1, it is characterized in that the data synchronization processing mode of real time data synchronization processing unit;

(1) adopt analog circuit phase-correction filter to carry out preliminary synchronisation to data waveform, analog circuit phase-correction filter has the characteristic with the complementation of simulated modulation filter, group delay is corrected to close to linear curve;

(2) data after preliminary synchronisation process will in the following way, processes:

Acceleration passage is resampled to 25.6KHz after carrying out continuous 10KHz low pass linear filtering;

Displacement passage is resampled to 5.12KHz after carrying out continuous 2KHz low pass linear filtering;

Speed channels is resampled to 2.56KHz after carrying out continuous 1KHz low pass linear filtering, then enters integrating circuit and complete integration, and then carries out phasing by the phase correction module of a digital circuit, offsets the phase deviation of different frequency;

Stress, strained channel are resampled to 2.56KHz after carrying out continuous 1KHz low pass linear filtering;

Traffic channel is resampled to 1KHz after carrying out continuous 500Hz low pass linear filtering;

Pressure channel is resampled to 1KHz after carrying out continuous 500Hz low pass linear filtering.