CN101474762A - Electric spark clearance discharge condition detection apparatus and method based on wavelet transformation - Google Patents
Electric spark clearance discharge condition detection apparatus and method based on wavelet transformation Download PDFInfo
- Publication number
- CN101474762A CN101474762A CNA2009100450500A CN200910045050A CN101474762A CN 101474762 A CN101474762 A CN 101474762A CN A2009100450500 A CNA2009100450500 A CN A2009100450500A CN 200910045050 A CN200910045050 A CN 200910045050A CN 101474762 A CN101474762 A CN 101474762A
- Authority
- CN
- China
- Prior art keywords
- discharge condition
- analog
- wavelet transformation
- conversion module
- digital conversion
- Prior art date
- Legal status (The legal status 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 status listed.)
- Granted
Links
Images
Abstract
The present invention relates to an electric spark clearance discharge condition detector which is in the special processing field and is based on wavelet transformation and a method thereof, wherein an analog optical coupler is connected with an electric spark machine tool through an inter-pole voltage signal outgoing cable. An analog-to-digital conversion module is connected at the back end of analog optical coupler. A DSP is connected at the back end of analog-to-digital conversion module. A CPLD is connected with the DSP, the USB interface and an SDRAM. A FLASH memory is connected with the DSP. The USB interface is connected with an upper machine through USB cable. The analog-to-digital conversion module samples the inter-pole voltage signal in electric spark machining and executes wavelet transformation procession to the signal in DSP. The wavelet coefficient comprising discharge condition information is obtained. The discharge condition of prior electric spark machining is obtained through the classifying decision and statistics summation of maximal value and minimum value of wavelet coefficient. The discharge condition of prior electric spark machining is transmitted to an upper machine. The electric spark clearance discharge condition detector of the invention has the advantages of excellent real-time property, excellent stability, excellent reliability, excellent accuracy and wide application sphere.
Description
Technical field
The present invention relates to the checkout gear and the method in a kind of special processing technology field, specifically, is a kind of electric spark clearance discharge condition detection apparatus and method based on wavelet transformation.
Background technology
Spark machined is come the material of ablation surface of the work by the Pulse Electric spark discharge between tool-electrode and workpiece, and electrode shape oppositely copies on the workpiece the most at last, this processing method has and is not subjected to workpiece material intensity, the restriction of mechanical performances such as hardness and do not have advantages such as macroscopical cutting force, be particularly suitable for the processing of difficult-to-machine material and complex-shaped workpieces, have a wide range of applications in industries such as Aeronautics and Astronautics, automobile, electronics, moulds.
Voltage across poles signal reaction in the spark machined complex environment between the electrode in the process, closely related with the efficient and the precision of spark machined.Usually, the discharge condition in the time of with spark machined is divided into four types: spark discharge, arc discharge, short circuit and open circuit.Load high pressure from interpolar and to puncturing normal spark discharge takes place, a bigger puncture time-delay is all arranged usually, arc discharge does not then generally puncture time-delay, or it is very little to puncture time-delay.Therefore,, the puncture time-delay can judge that puncture time-delay detection method is a kind of important method in the spark discharge state-detection to the discharge condition of spark discharge by being carried out timing.But because the complexity of spark machined, present existing interpolar condition detection method can be subjected to the interference of various noises and wave distortion in the discharge process usually, has limited the accuracy and the reliability of testing result.Therefore, Chinese scholars is making great efforts to seek new spark discharge condition detection method always.
Wavelet analysis is a kind of Time-Frequency Analysis Method, with respect to traditional frequency-domain analysis method, more be applicable to signal average and the continuous non-stationary signal that changes of variance are analyzed, characteristics with multiresolution analysis, can observe the local property in time domain and the frequency domain simultaneously, and the result of Fourier transformation can lose the information on the time domain.The voltage signal that produces in the time of spark machined is exactly a kind of non-stationary signal with strong nonlinearity, and this class signal is well suited for analyzing with method of wavelet, the information relevant with the interpolar discharge state that draws in the signal to be contained.
Find through literature search prior art, at periodical " The International Journal ofAdvanced Manufacturing Technology " (international advanced manufacturing technology magazine) volume17, pp339-343,2001, in " Waveform Monitoring of Electric Discharge Machining byWavelet Transform " (adopting the spark discharge waveforms detection of wavelet transformation), by the frequency component of certain high frequency band among the voltage and current wavelet transformation result is analyzed, judge the spark discharge state, discharge condition is divided into open circuit, regular picture, short circuit, between electric arc and arteries and veins five types.But this article does not have detailed argumentation to the process of how to carry out wavelet transformation and how to judge according to transformation results.And the wavelet transformation result that it draws falls within a certain than in the narrow-band, if will realize accurate judgement to discharge condition, must suppose and guarantee can not exist in regional other zone in addition of voltage and current sudden change the component in this frequency band.And because the complexity and the randomness of spark machined, this hypothesis can not be guaranteed, and also there is a certain distance by the transformation results that comparatively Utopian discharging model draws in this apart from practical application.
Summary of the invention
The objective of the invention is at the deficiencies in the prior art, a kind of electric spark clearance discharge condition detection apparatus and method based on wavelet transformation is provided, adopt the low frequency wavelet coefficient as the foundation of judging discharge condition, it is good to have real-time, reliable and stable, electric spark clearance discharge condition accurate and applied widely detects.
The present invention is achieved by the following technical solutions:
Electric spark clearance discharge condition detection apparatus based on wavelet transformation involved in the present invention comprises checkout gear body, host computer and voltage across poles signal outgoing cable.The checkout gear body comprises simulation optocoupler, analog-to-digital conversion module, DSP (digital signal processor), CPLD (CPLD), USB (USB) interface, SDRAM (synchronous DRAM), FLASH memory, USB cable.Connected mode is: the simulation optocoupler links to each other with electric spark machine tool by voltage across poles signal outgoing cable, analog-to-digital conversion module is connected the rear end of simulation optocoupler, DSP is connected the rear end of analog-to-digital conversion module, CPLD links to each other with DSP, USB interface, SDRAM, the FLASH memory links to each other with DSP, and USB interface links to each other with host computer by USB cable.
The spark machined voltage signal that produces at electric spark machine tool is input to the simulation optocoupler by voltage across poles signal outgoing cable, analog voltage signal after the simulation light-coupled isolation, convert data signal to by analog-to-digital conversion module, be input among the DSP, primary signal is carried out wavelet transformation to DSP and discharge condition is judged, try to achieve average discharge condition coefficient,, be input to host computer by USB cable through CPLD and USB interface.SDRAM provides more memory space for system, when the data buffer area among the DSP is crowded, and can partial data is temporary in SDRAM.Stored the DSP program in the FLASH memory, after the checkout gear body powered on, DSP moved from FLASH memory read program fetch automatically.
Described analog-to-digital conversion module is meant, switching rate more than 1Msps, sampling precision is greater than 10 analog-to-digital conversion module, make to enter the signal that DSP handles and have higher precision.
The floating-point operation ability of described DSP reaches more than the 1.6TFlops, and the sampled data that enters is accomplished real-time sampling, handles in real time, in real time output.
Electric spark clearance discharge condition detection method based on wavelet transformation involved in the present invention, be the voltage across poles signal in electrical spark working man-hour to be sampled by analog-to-digital conversion module, and after the signal that in DSP (digital signal processor) sampling is obtained carries out wavelet transform process, draw the wavelet coefficient that comprises discharge condition information; By wavelet coefficient maximum and minimizing classification being judged and the statistics summation, draw current spark machined discharge condition, and send host computer to, for the operation of control system provides foundation through USB (USB) interface.
Said method of the present invention comprises the steps:
1. with after the electrical spark working voltage signal usefulness simulation light-coupled isolation in man-hour, insert analog-to-digital conversion module and carry out analog-to-digital conversion.
Described simulation optocoupler has the above bandwidth of DC～1MHz, be used for the complete electrical isolation between electric spark machine tool and the checkout gear body, the various information that guarantee the spark discharge voltage signal can not lost before analog-to-digital conversion module carries out analog-to-digital conversion entering.
2. analog-to-digital conversion module is to analog signal sampling, and analog-to-digital conversion module is connected with DSP, and DMA (direct memory visit) module of the data that obtain after the sampling by DSP is deposited in the memory of DSP;
In this step, analog-to-digital precision is 12, and sample rate is 1MHz, promptly per 1 μ s once sampling, and for most of spark machined, the sample rate of this speed is enough reflecting voltage waveforms accurately.
3. behind 1024 sampled points of storage, adopt Mallat algorithm (the special algorithm of horse traction),, these 1024 raw data points are carried out 4 layers of wavelet transformation, obtain the low frequency wavelet coefficient of 64 points with Daubechies (channel ratio is strange) three rank wavelet functions.
3. 2. step adopt the PING-PONG pattern with step, promptly when 3. step is handled the initial data of certain segment memory, 2. step writes data to another segment memory, treat that 3. step finish after the processing of this segment data, then the new data segment that 2. step is just now write is handled, and that section that 3. 2. step then handled step just now left with the memory of initial data and write new data.Because the utilization of dma module can read and write memory space under the situation that does not consume CPU (central processing unit) resource, has made full use of the resource of DSP, has also guaranteed the real-time of system.
4. because the processing speed of DSP is very fast, behind four layers of wavelet transformation finishing 1024 data points, 1024 sampled points of next period are not also gathered and are finished, therefore, CPU has time enough that 64 low frequency coefficients that obtain are further processed, and promptly CPU carries out local maximum and local minimizing judgement and analysis to 64 low frequency coefficients that obtain.At first, seek the local maximum and the local minimum of these 64 coefficients, after whenever finding a local maximum or local minimum, utilize previous preset threshold that these extreme values are sorted out.Local maximum then is considered as taking place arc discharge less than a certain threshold value, and local minimum then is considered as failing disruptive discharge greater than a certain threshold value, is the pulse of an open circuit.
5. DSP draws the discharge condition coefficient of this section primary signal by the statistics to local extreme value analysis result.
Particularly, will sue for peace separately, calculate regular picture pulse shared ratio in these pulses, the discharge condition coefficient in the time of the 1ms that draws at these 1024 sampled point representatives through the umber of pulse of sorting out
6. 1. above-mentioned～5. five steps after 10 times, DSP averages to 10 discharge condition coefficients of this accumulation, draws the average discharge condition coefficient in this period repeatedly.
7. DSP is by CPLD (CPLD), and USB interface and USB cable be the average discharge condition coefficient input host computer that obtains, for the operation of control system provides foundation.
The present invention delays time as one of foundation of judging puncturing, and it is more little to puncture time-delay, and then corresponding local maximum is more little.Puncturing time-delay is the important information of reflection chip removal situation, and it is big to puncture time-delay, shows that then discharging gap is big, and the chip removal condition is good, if it is too little to puncture time-delay, then the physical state of contrasted between solid dielectric worsens (the galvanic corrosion production concentration is excessive or local temperature is too high).For the comparatively difficult process of chip removals such as titanium alloy processing, with the index of puncture time-delay as judgement, and list is significant with the spark discharge ratio as index.
It is a kind of detection method of unit that the present invention is actually with the individual pulse, it is neither the calculating that one section voltage data is averaged, do not probe into the discharge condition of certain the concrete point in each pulse yet, and judge and detect by the distribution situation of adding up a large amount of points.It neither carry out timing to the open-circuit voltage section of certain pulse, determines that it punctures the length of time-delay.It is to be calculating object with a plurality of pulses, with the individual pulse is to judge unit, judges discharged condition according to the distribution of local peaking behind the wavelet transformation.Owing to be to be unit with the individual pulse, just can not be subjected to the wide influence that waits processing rule standard to change of intercadence, only need to adjust where necessary the threshold value when sorting out.It can also pass through the analysis to the relation of adjacent posivtive spike negative peak simultaneously, obtains some other information, such as whether carbon deposit etc. being arranged, also can combine analysis with electric current or high fdrequency component detection.
Wavelet analysis can reduce the data point of initial data section, the data of 1024 points through four layers of wavelet transformation after, tapering to only has 64 points, thereby has formed a simple figure, for the statistics of Local Extremum has been created condition.This simple graphics class is similar to the fuzzy Judgment of people to the waveform overall impression, because the people to the judgement of spark discharge waveform, can not remove to seek the electric current of certain sampled point, voltage, the situation of high fdrequency component, but with a kind of comparatively fuzzy rule, from the angle of macroscopic view, make whole evaluation.
Because the present invention surpasses the high-speed AD converter of 1Msps by high sampling rate and the embedded system that DSP builds realizes, therefore have strict sequential and higher stable degree, be not vulnerable to interference and influence that environment for use changes.And, because its most of function is not to solidify with the form of digital circuit, but be kept in the memory with the form of program, therefore have suitable flexibility again, can make corresponding modification according to different situations.
The present invention also can satisfy the requirement of real-time fully, and this is by the decision of the characteristics of dsp system.DSP has higher clock frequency, the Harvard structure that adopts data/address bus to separate with program bus, and the mode of employing pile line operation and parallel computation, a large amount of taking advantage of adds processing in the energy fast processing wavelet transformation process.The characteristics of multiregister help handling a plurality of data simultaneously, have reduced the number of transmissions of data between CPU and on-chip memory.The use of dma module can directly not deposit into on-chip memory to the data that ADC sends by CPU, realizes the calculating of sampling edge limit, can realize the communication between main frame and the DSP simultaneously yet.According to existing software emulation result, by optimizing code, give full play to the advantage of pipeline processes and parallel computation, 1024 points of 1ms in the time are carried out four layers of wavelet transformation of Daubechies3 rank small echo, only need more than 3000 clock cycle to finish.Promptly, when the DSP clock frequency is 200MHz, only need just can finish wavelet transformation one time less than the time of 0.02ms, even added the time of wavelet transformation result treatment, consider storage interference and other some expenses again, also can finish the analyzing and processing of a data block fully at 1ms in the time, and not have loss of data between data block and the data block, satisfy the requirement of handling in real time.When The results of data processing is sent to control system, the delay of 1ms is arranged approximately.On the one hand, this delay is inevitably, because any detection method all can produce delay; On the other hand, this delay also is negligible, because the common every 10ms of control system is just to checkout gear inquiry single step of releasing electricity condition information, and will be suitable for control law with the result after the value weighting repeatedly, and the information of 1ms is normally inappreciable.
Description of drawings
Fig. 1 is a spark discharge state waveform schematic diagram;
Wherein: spark discharge (a) has bigger puncture time-delay, it is a good spark discharge pulse, and spark discharge (b) is although also there is the time-delay of puncturing, and it is less to puncture time-delay, and the interpolar state that shows spark discharge (b) enters unsettled discharge than spark discharge (a) is easier.
Fig. 2 is the original waveform figure among the embodiment 1.
Fig. 3 is the low frequency wavelet coefficient figure behind wavelet transformation among the embodiment 1.
Fig. 4 and Fig. 5 are original waveform figure among the embodiment 2 and the low frequency wavelet coefficient figure behind wavelet transformation.
Fig. 6 and Fig. 7 are original waveform figure among the embodiment 3 and the low frequency wavelet coefficient figure behind wavelet transformation.
Fig. 8 is the flow chart that calculates the discharge condition coefficient by wavelet coefficient.
Fig. 9 is based on the spark discharge condition checkout gear block diagram of wavelet transformation;
Wherein: checkout gear body 1, host computer 2, simulation optocoupler 3, analog-to-digital conversion module 4, DSP5, CPLD6, USB interface 7, SDRAM8, FLASH memory 9 and voltage across poles signal outgoing cable 10, USB cable 11, electric spark machine tool 12.
The specific embodiment
Below in conjunction with accompanying drawing embodiments of the invention are elaborated: present embodiment is being to implement under the prerequisite with the technical solution of the present invention, provided detailed embodiment and concrete operating process, but protection scope of the present invention is not limited to following embodiment.
As shown in Figure 1, the discharge condition in the time of spark machined is divided into four types: spark discharge, arc discharge, short circuit and open circuit.In spark discharge (a) pulse, the long puncture time-delay of a tangible time is arranged, show that interpolar is in good condition, the processing bits that spark discharge produces can successfully be discharged region of discharge; And the puncture of spark discharge (b) pulse time-delay is short than spark discharge (a) pulse, but still belongs to normal range (NR).In the electric arc pulse, when being elevated to open-circuit voltage as yet, voltage just do not discharge, almost puncture time-delay, represent that at this moment interpolar state has the trend of deterioration, the stability and the efficient of spark machined is had a negative impact.Short circuit and open circuit pulse then are the extreme cases of anode-cathode distance.By detection method involved among the present invention and device, can exactly the interpolar pulse be divided into above-mentioned spark discharge, arc discharge, short circuit and open circuit four classes, the trend that perhaps detects the inclined to one side short circuit of current interpolar state or open a way partially.
As shown in Figure 9, the electric spark clearance discharge condition detection apparatus based on wavelet transformation that present embodiment relates to comprises: checkout gear body 1, host computer 2 and voltage across poles signal outgoing cable 10.Checkout gear body 1 comprises simulation optocoupler 3, analog-to-digital conversion module 4, and DSP 5, and CPLD 6, USB interface 7, SDRAM 8, FLASH memory 9, USB cable 11.Connected mode is: simulation optocoupler 3 links to each other with electric spark machine tool 12 by voltage across poles signal outgoing cable 10, analog-to-digital conversion module 4 is connected the rear end of simulation optocoupler 3, DSP5 is connected the rear end of analog-to-digital conversion module 4, CPLD 6 and DSP5, USB interface 7, SDRAM 8 links to each other, and FLASH memory 9 links to each other with DSP5, and USB interface 7 links to each other with host computer 2 by USB cable 11.
The spark machined voltage signal that produces at electric spark machine tool 12 is input to simulation optocoupler 3 by voltage across poles signal outgoing cable 10, analog voltage signal after simulation optocoupler 3 is isolated, be converted to data signal by analog-to-digital conversion module 4, be input among the DSP 5,5 pairs of primary signals of DSP carry out wavelet transformation and discharge condition is judged, try to achieve average discharge condition coefficient,, be input to host computer 2 by USB cable 11 through CPLD 6 and USB interface 7.SDRAM 8 provides more memory space for system, when the data buffer area among the DSP 5 is crowded, and can partial data is temporary in SDRAM 8.Stored DSP 5 programs in the FLASH memory 9, after checkout gear body 1 powered on, DSP 5 moved from 9 fetch programs of FLASH memory automatically.
Analog-to-digital conversion module 4 has the above switching rate of 1Msps, and sampling precision is greater than 10, makes to enter the signal that DSP 5 handles and have higher precision.
The floating-point operation ability of DSP 5 reaches more than the 1.6TFlops, and the sampled data that enters is accomplished to implement sampling, handles in real time, in real time output.
The electric spark clearance discharge condition detection method that present embodiment relates to based on wavelet transformation, by the high speed analog-to-digital conversion chip voltage across poles signal in electrical spark working man-hour is sampled, and after the signal that in DSP sampling is obtained carries out wavelet transform process, draw the wavelet coefficient that comprises discharge condition information.By wavelet coefficient maximum and minimizing classification are judged and the statistics summation, draw current spark machined discharge condition, and send host computer to through USB interface.
Concrete implementation step is:
1. the voltage signal in electrical spark working man-hour enters analog-to-digital conversion module 4 and carries out analog-to-digital conversion after isolating through simulation optocoupler 3.Simulation optocoupler 3 has the above bandwidth of DC～1MHz, be used for the complete electrical isolation between electric spark machine tool 12 and the checkout gear body 1, the various information that guarantee the spark discharge voltage signal can not lost before analog-to-digital conversion module 4 carries out analog-to-digital conversion entering.
2. 4 pairs of analog signal samplings of analog-to-digital conversion module, analog-to-digital conversion module is connected with DSP 5, and the data that obtain after the sampling dma module by DSP 5 is deposited in the memory of DSP 5.
3. DSP 5 adopts the special algorithm of horse traction behind 1024 sampled points of storage, with the strange three rank wavelet functions of channel ratio, these 1024 raw data points is carried out 4 layers of wavelet transformation, obtains the low frequency wavelet coefficient of 64 points;
4. behind four layers of wavelet transformation finishing 1024 data points, 5 pairs of these wavelet coefficients of DSP carry out local maximum and local minimizing judgement and analysis.
5. DSP 5 draws the discharge condition coefficient of this section primary signal by the statistics to local extreme value analysis result.
6. after above-mentioned repeatedly five steps 10 of DSP 5 time, DSP 5 averages to 10 discharge condition coefficients of this accumulation, draws average discharge condition coefficient in certain time period.
7. DSP 5 is by CPLD 6, and USB interface 7 and USB cable 11 be the average discharge condition coefficient input host computer 2 that obtains, for the operation of control system provides foundation
2. step has used the dma module among the DSP 5, does not take cpu resource.Carrying out step 2. the time, DSP5 can carry out step simultaneously and 3.～7. wait every processing.Step 2. in, the sampling precision of analog-to-digital conversion module is 12, sample rate is 1MHz, promptly per 1 μ s once samples.
3. 2. step adopt the PING-PONG pattern with step, promptly when 3. step is handled the initial data of certain segment memory, 2. step writes data to another segment memory, treat that 3. step finish after the processing of this segment data, then the new data segment that 2. step is just now write is handled, and that section that 3. 2. step then handled step just now left with the memory of initial data and write new data.
3. step adopts the Mallat algorithm, with Daubechies3 rank wavelet function, 1024 raw data points is carried out 4 layers of wavelet transformation, obtains the low frequency wavelet coefficient of 64 points.
Step is 4. after drawing local maximum and local minimum, according to the threshold value of setting these extreme values are sorted out, less local maximum is considered as not puncturing the arc discharge or the short circuit pulse of time-delay, bigger local minimum is considered as not having the open circuit pulse that punctures
Fig. 8 describe in detail step 4. to step 5. in, the result analyzes to wavelet transformation, and finally draws the flow process of discharge condition coefficient.After obtaining wavelet conversion coefficient, seek the step of Local Extremum, concrete grammar is the size of more adjacent three data points, if that middle data point is greater than two adjacent data points about it, then be considered as local maximum, if less than two adjacent data points of the left and right sides, then be considered as local minimum.After finding Local Extremum, judge if maximum, and its value is less than local maximum threshold value MAXth, then short circuit pulse is counted Ps and is added 1 partially; If minimum, and its value is greater than local minimum threshold value MINth, and the umber of pulse Po that then opens a way partially adds 1.Whenever find a local minimum, then pulse sum Pa adds 1.The circulation said process all passes through judgement up to all wavelet conversion coefficients, after Local Extremum is sought and finished, calculates discharge condition coefficient c, discharge condition coefficient c=(Pa+Po-Ps)/Pa.Discharge pulse coefficient c approaches 1 more, shows that then current discharge condition is good more; Discharge pulse coefficient c approaches 0 more, shows that then current discharge condition approaches short circuit more; Discharge pulse coefficient c approaches 2 more, shows that then current discharge condition approaches open circuit more.Thereby, by different discharge condition coefficients, can instruct control system to make different feeding rollback motions.Above-mentioned local maximum threshold value MAXth and local minimum threshold value MINth can set according to actual conditions, for example, if local maximum threshold value MAXth is made as 250, MINth is made as 350 with the local minimum threshold value, then, can calculate discharge condition coefficient c according to wavelet conversion coefficient according to the flow process among Fig. 8.
Below three embodiment, be to from step 3. to step application implementation 5., promptly from step primary signal carried out wavelet transformation draw the wavelet transformation result 3., to step classification 4., to the step discharge condition coefficient that calculates this section primary signal 5. to Local Extremum.
Shown in Fig. 2,3, in the oscillogram of the primary signal s of Fig. 2, the A place has a pulse to fail puncture to form discharge, " open circuit " pulse promptly occurred, and the 4th layer of wavelet transformation among Fig. 3 corresponding with it as a result the low frequency coefficient of CA4 present a very big local minimum.The B place has two pulses to fail to be flushed to just discharge in advance of breakdown voltage among Fig. 2, does not exist to puncture time-delay, and the situation that has arcing exists, and this presents two minimum local maximums in Fig. 3.The C place has one to puncture the minimum pulse of time-delay among Fig. 2, and a corresponding less local maximum is also arranged in Fig. 3.Two of the D place pulses have bigger puncture time-delay among Fig. 2, are discharge pulses preferably, and they are corresponding two bigger local maximums in Fig. 3.As seen, can effectively judge the situation of each pulse by the wavelet transformation result.
If local maximum threshold value MAXth is made as 250, local minimum threshold value MINth is made as 350, then according to the flow process among Fig. 8, can draw the discharge condition coefficient of primary signal shown in Figure 2.
Pulse sum Pa=22
Umber of pulse Po=1 partially opens a way
Short circuit pulse is counted Ps=8 partially
Discharge condition coefficient c=0.68.
As shown in Figure 4 and Figure 5, local maximum threshold value MAXth=250 wherein, local minimum threshold value MINth=350, pulse sum Pa=20, the umber of pulse of opening a way partially Po=1, short circuit pulse is counted Ps=12, discharge condition coefficient c=0.45 partially.
As shown in Figure 6 and Figure 7, local maximum threshold value MAXth=250, local minimum threshold value MINth=350, pulse sum Pa=20, the umber of pulse of opening a way partially Po=8, short circuit pulse is counted Ps=6, discharge condition coefficient c=1.10 partially.
Claims (8)
1, a kind of electric spark clearance discharge condition detection apparatus based on wavelet transformation, it is characterized in that comprising the checkout gear body, host computer and voltage across poles signal outgoing cable, described checkout gear body comprises the simulation optocoupler, analog-to-digital conversion module, digital signal processor, CPLD, USB interface, synchronous DRAM, the FLASH memory, USB cable, wherein: the simulation optocoupler links to each other with electric spark machine tool by voltage across poles signal outgoing cable, analog-to-digital conversion module is connected the rear end of simulation optocoupler, digital signal processor is connected the rear end of analog-to-digital conversion module, CPLD and digital signal processor, USB interface, synchronous DRAM links to each other, the FLASH memory links to each other with digital signal processor, and USB interface links to each other with host computer by USB cable; The spark machined voltage signal that electric spark machine tool produces is input to the simulation optocoupler by voltage across poles signal outgoing cable, analog voltage signal after the simulation light-coupled isolation, convert data signal to by analog-to-digital conversion module, be input in the digital signal processor, primary signal is carried out wavelet transformation with digital signal processor and discharge condition is judged, try to achieve average discharge condition coefficient, through CPLD and USB interface, be input to host computer by USB cable, synchronous DRAM provides more memory space for system, when the data buffer area in the digital signal processor is crowded, can partial data is temporary in synchronous DRAM, stored digital signal processor program in the FLASH memory, after the checkout gear body powered on, digital signal processor moved from FLASH memory read program fetch automatically.
2, the electric spark clearance discharge condition detection apparatus based on wavelet transformation according to claim 1 is characterized in that, described analog-to-digital conversion module is meant, switching rate more than 1Msps, sampling precision is greater than 10 analog-to-digital conversion module.
3, a kind of electric spark clearance discharge condition detection method based on wavelet transformation is characterized in that, comprises the steps:
1. with after the electrical spark working voltage signal usefulness simulation light-coupled isolation in man-hour, insert analog-to-digital conversion module and carry out analog-to-digital conversion;
2. analog-to-digital conversion module is to analog signal sampling, and analog-to-digital conversion module is connected with digital signal processor, and the direct memory access modules of the data that obtain after the sampling by digital signal processor is deposited in the memory of digital signal processor;
3. behind 1024 sampled points of storage, adopt the special algorithm of horse traction,, these 1024 raw data points are carried out 4 layers of wavelet transformation, obtain the low frequency wavelet coefficient of 64 points with the strange three rank wavelet functions of channel ratio;
4. behind four layers of wavelet transformation finishing 1024 data points, CPU carries out local maximum and local minimizing judgement and analysis to 64 low frequency coefficients that obtain;
5. digital signal processor is by adding up the discharge condition coefficient that draws this section primary signal to local extreme value analysis result;
6. 1. above-mentioned～5. five steps after 10 times, digital signal processor is averaged to 10 discharge condition coefficients of this accumulation, draws the average discharge condition coefficient in this period repeatedly;
7. digital signal processor by CPLD, USB interface and USB cable the average discharge condition coefficient input host computer that obtains, for the operation of control system provides foundation.
4, the electric spark clearance discharge condition detection apparatus based on wavelet transformation according to claim 3 is characterized in that, described simulation optocoupler has the above bandwidth of DC～1MHz.
5, the spark discharge condition detection method based on wavelet transformation according to claim 3 is characterized in that, step 2. in, analog-to-digital precision is 12, sample rate is 1MHz.
6, the spark discharge condition detection method based on wavelet transformation according to claim 3, it is characterized in that, 3. 2. step adopt the PING-PONG pattern with step, promptly when 3. step is handled the initial data of certain segment memory, 2. step writes data to another segment memory, treat that 3. step finish after the processing of this segment data, then the new data segment that 2. step is just now write is handled, and that section that 3. 2. step then handled step just now left with the memory of initial data and write new data.
7, the spark discharge condition detection method based on wavelet transformation according to claim 3, it is characterized in that, step 4. in, at first, seek the local maximum and the local minimum of these 64 coefficients, after whenever finding a local maximum or local minimum, utilize previous preset threshold that these extreme values are sorted out, local maximum is less than a certain threshold value, then be considered as taking place arc discharge, local minimum then is considered as failing disruptive discharge greater than a certain threshold value, is the pulse of an open circuit.
8, the spark discharge condition detection method based on wavelet transformation according to claim 3, it is characterized in that, step 5. in, to sue for peace separately through the umber of pulse of sorting out, calculate regular picture pulse shared ratio in these pulses, the discharge condition coefficient in the time of the 1ms that draws at these 1024 sampled point representatives.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN200910045050A CN100595030C (en) | 2009-01-08 | 2009-01-08 | Electric spark clearance discharge condition detection apparatus and method based on wavelet transformation |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN200910045050A CN100595030C (en) | 2009-01-08 | 2009-01-08 | Electric spark clearance discharge condition detection apparatus and method based on wavelet transformation |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN101474762A true CN101474762A (en) | 2009-07-08 |
| CN100595030C CN100595030C (en) | 2010-03-24 |
Family
ID=40835583
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN200910045050A Active CN100595030C (en) | 2009-01-08 | 2009-01-08 | Electric spark clearance discharge condition detection apparatus and method based on wavelet transformation |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN100595030C (en) |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105522238A (en) * | 2014-10-28 | 2016-04-27 | 佛山科学技术学院 | Discharging gap state detection module based on pulse sequence analysis |
| CN106735636A (en) * | 2017-03-31 | 2017-05-31 | 山东豪迈机械科技股份有限公司 | Electrical discharge machining discharging headlamp circuit and method |
| CN107891199A (en) * | 2017-12-01 | 2018-04-10 | 中州大学 | Spark discharge condition checkout gear and recognition methods |
| CN108436203A (en) * | 2018-04-28 | 2018-08-24 | 南通伊阳精密机械有限公司 | Adaptive wire-electrode cutting impulsing power source and the servo tracking control device for monitoring the power supply |
| CN110142471A (en) * | 2019-07-02 | 2019-08-20 | 哈尔滨工业大学 | Insulative ceramic coatings-metal electrical discharge machining paradoxical discharge condition checkout gear and method |
| CN112393276A (en) * | 2019-08-13 | 2021-02-23 | 广东百威电子有限公司 | Pulse ignition control method for gas appliance |
-
2009
- 2009-01-08 CN CN200910045050A patent/CN100595030C/en active Active
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105522238A (en) * | 2014-10-28 | 2016-04-27 | 佛山科学技术学院 | Discharging gap state detection module based on pulse sequence analysis |
| CN105522238B (en) * | 2014-10-28 | 2017-09-22 | 佛山科学技术学院 | Discharging gap-state detection module based on pulse-train analysis |
| CN106735636A (en) * | 2017-03-31 | 2017-05-31 | 山东豪迈机械科技股份有限公司 | Electrical discharge machining discharging headlamp circuit and method |
| CN106735636B (en) * | 2017-03-31 | 2019-11-19 | 山东豪迈机械科技股份有限公司 | Electrical discharge machining discharging headlamp circuit and method |
| CN107891199A (en) * | 2017-12-01 | 2018-04-10 | 中州大学 | Spark discharge condition checkout gear and recognition methods |
| CN108436203A (en) * | 2018-04-28 | 2018-08-24 | 南通伊阳精密机械有限公司 | Adaptive wire-electrode cutting impulsing power source and the servo tracking control device for monitoring the power supply |
| CN110142471A (en) * | 2019-07-02 | 2019-08-20 | 哈尔滨工业大学 | Insulative ceramic coatings-metal electrical discharge machining paradoxical discharge condition checkout gear and method |
| CN112393276A (en) * | 2019-08-13 | 2021-02-23 | 广东百威电子有限公司 | Pulse ignition control method for gas appliance |
Also Published As
| Publication number | Publication date |
|---|---|
| CN100595030C (en) | 2010-03-24 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN100595030C (en) | Electric spark clearance discharge condition detection apparatus and method based on wavelet transformation | |
| CN106443335A (en) | Lightning stroke fault identification method and system | |
| CN108470163B (en) | Rail turnout plate gap defect detection method and terminal equipment | |
| US20170296081A1 (en) | Frame based spike detection module | |
| CN104459644A (en) | Self-adaptive constant false alarm detecting method used for detecting radar video signals | |
| Williams et al. | Helicopter transmission fault detection via time-frequency, scale and spectral methods | |
| CN108270543A (en) | A kind of side-channel attack preprocess method based on small echo spatial domain correlation method | |
| CN113484700A (en) | Switch cabinet partial discharge detection method based on indoor intelligent inspection robot | |
| CN103166606B (en) | The method and apparatus that digital pulse signal is screened | |
| WO2019075913A1 (en) | Signal processing system applied to remove otdr noise | |
| CN104749532B (en) | A kind of spacecraft power supply system failure detection method and device | |
| CN107398612A (en) | A kind of electric spark clearance discharge condition detection apparatus and method based on cluster analysis | |
| CN108594156B (en) | Improved current transformer saturation characteristic identification method | |
| CN114367710B (en) | Electric spark machining control method based on deep learning and acoustic emission signals | |
| CN116718874A (en) | Traveling wave fault positioning method, system and device | |
| CN208147865U (en) | A kind of sound recognition system applied to robot used for intelligent substation patrol | |
| CN110600060B (en) | Hardware audio active detection HVAD system | |
| CN113188651B (en) | Vibration signal feature extraction and tool wear value prediction method based on IMF-PSD spectral line | |
| CN101691632B (en) | Method for rapidly judging peak | |
| CN114114400A (en) | Microseism event effective signal pickup method | |
| CN113848256A (en) | Real-time detection method for ultrasonic first-motion waves | |
| Gao et al. | Source number estimation based on improved singular value decomposition at low SNR | |
| CN100534384C (en) | Magnetocardiogram signals collecting processing method based on digital signal processing and device thereof | |
| CN111590390B (en) | Cutter wear state real-time assessment method and system, storage medium and terminal | |
| Chen et al. | Real-time detection and classification for targeted marine mammals |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant |