CN104856666A - Bio-electricity signal monitoring system based on LabVIEW - Google Patents

Abstract

The invention discloses a bio-electricity signal monitoring system based on LabVIEW. Multipath bio-electricity signals are extracted by bio-electricity signal extracting equipment, and are subjected to parallel processing, and people to be monitored are monitored by using a bio-electricity signal monitoring interface. During specific configuration, the people to be monitored can selectively call a library function pre-stored in the system through a signal processing interface according to own requirements to perform real-time processing on the bio-electricity signals, and finally the bio-electricity signals are displayed on a display frame on the signal processing interface, so that the monitoring on physiological indexes of the people to be monitored can be completed.

Description

A kind of bioelectrical signals monitoring system based on LabVIEW

Technical field

The invention belongs to physiology monitoring technical field, more specifically say, relate to a kind of bioelectrical signals monitoring system based on LabVIEW.

Background technology

Bioelectrical signals is the key character parameter of health state, also effectively reflects brain activity situation and the human body moving situation of people simultaneously.By to the analysis of human biological signal and monitoring, the health status of human body can be held exactly, more by the feature identification to different bioelectrical signals, the operation and controlling to external equipment can be realized.

Bioelectrical signals analysis in the market and monitoring system functional too single, signal processing algorithm is comparatively simple, and most homogeneous system program does not possess good portability, as the brain electric transducer of science and technology based on TGAM chip design is read based on the EEG signals extraction module of ADS1299 chip design and god by TI company, therefore, above two kinds of equipment are difficult to effectively promote in the market.

More specifically say, compared to the EEG signals extraction module of TI company based on ADS1299 chip design, it is little that the present invention has volume, multiple powering mode, and heat power consumption is little, carries the advantage such as bioelectrical signals extraction element and multi-purpose software display.Read the brain electric transducer of science and technology based on TGAM chip design compared to god, it is low that the present invention has power consumption, and communication is stable, the advantages such as multichannel collecting.

Summary of the invention

The object of the invention is to overcome the deficiencies in the prior art, a kind of bioelectrical signals monitoring system based on LabVIEW is provided, by signal processing algorithm, the bioelectrical signals collected is analyzed, realize the Real-Time Monitoring to body state.

For achieving the above object, the present invention is based on the bioelectrical signals monitoring system of LabVIEW, it is characterized in that, comprising:

One bioelectrical signals extraction equipment, comprise electroencephalogramsignal signal collection equipment and multiple limbs signal collecting device, in two collecting devices, include spring steel plate, stretchy nylon band and dry electrode slice, wherein, in electroencephalogramsignal signal collection equipment, at least comprise 3 dry electrode slices; Dry electrode slice, is stuck in during by wearing bioelectrical signals extraction equipment on head and limbs by solder attachment on the spring steel plate of correspondence, and welding position is corresponding with bioelectrical signals active regions; Stretchy nylon band is positioned on spring steel plate, can add the closeness of contact of capable and experienced electrode slice and human body further, increases the stability of acquiring biological electric signals; When extracting bioelectrical signals, bioelectrical signals extraction equipment is worn on head and each position of limbs, extract the bioelectrical signals on brain and each position of limbs respectively, then send to PCB circuit integrated module by the electromagnetic shield signal line that collecting device connects;

Many electromagnetic shield signal lines, adopt the anti-tampering design of multilamellar, wherein internal layer is signal transmssion line, and skin is rubber layer, is tinsel envelope in the periphery of signal transmssion line;

One PCB circuit integrated module, comprises electromagnetic interface filter, preamplifier and analog-digital converter; For receiving the multichannel bioelectrical signals that bioelectrical signals extraction equipment gathers, and by the bioelectrical signals parallel processing of multichannel, then send to zigbee group-net communication module;

After PCB circuit integrated module receives bioelectrical signals, first pass through the electromagnetic noise that the outer frequency of electromagnetic interface filter filter out-band is higher, filtered bioelectrical signals is input to analog-digital converter after preamplifier amplifies, the bioelectrical signals of simulation is converted into the digital signal being easy to receive and dispatch by analog-digital converter again, and sends to zigbee group-net communication module;

One zigbee group-net communication module contains multiple COM1, can walk abreast and receive the multichannel bioelectrical signals of PCB circuit integrated module transmission, then be transmitted to bioelectrical signals observation interface, realizes the network type communication that many receipts are multiple;

One signal process function storehouse, signal processing interface, by function in call signal process function library, processes bioelectrical signals, re-sends to bioelectrical signals observation interface;

One bioelectrical signals observation interface, by the tab control on bioelectrical signals observation interface, selects entering signal observation interface or signal processing interface;

On signal monitoring interface, COM1 between zigbee group-net communication module and bioelectrical signals observation interface can be set, select signalling channel when needing entering signal process interface, when communicating, the signal quality of bioelectrical signals directly can be observed on instrumental panel, and the relevant information of bioelectrical signals demonstrates the bioelectrical signals on brain and limbs respectively by 2-D display box;

On signal processing interface, user can function as required in selectivity call signal process function library, then arranges according to the function called, and as Choose filtering function, then sets out and leaches frequency range; As selected trap function, then set out the Hz noise frequency range that trap function is removed; As selected wavelet function, then set the female ripple type of small echo, select the number of plies of wavelet decomposition; As selected neural network function, then when calling this function, train neural network classifier parameter; Bioelectrical signals is after above-mentioned process completes, and its relevant information demonstrates the bioelectrical signals on brain and limbs respectively by 2-D display box.

Goal of the invention of the present invention is achieved in that

The present invention is based on the bioelectrical signals monitoring system of LabVIEW, extracted the bioelectrical signals of multichannel by bioelectrical signals extraction equipment, then the bioelectrical signals of parallel processing multichannel, utilize bioelectrical signals observation interface to monitor monitored person.In concrete configuration, monitored person according to the needs of self, by the built-in function prestored in signal processing interface selective calling system, can process bioelectrical signals in real time, last again by the display box display on signal processing interface, complete the monitoring to monitored person's physical signs.

Meanwhile, the bioelectrical signals monitoring system that the present invention is based on LabVIEW also has following beneficial effect:

(1), in the present invention, bioelectrical signals extraction equipment forms based on spring steel, stretchy nylon band and dry electrode slice.The modes such as existing like product many employings electrode cap, adhesive patches extract bioelectrical signals, and compared with electrode cap, the present invention has that quality is light, volume is little, low cost and other advantages; Compared with adhesive patches, the present invention has and can repeatedly use, adjustable with human body bonding tightness, the advantage such as conveniently to dress and take off;

(2), by bioelectrical signals observation interface can observe the bioelectrical signals of multiple passage simultaneously, also manually can set communication interface, simultaneously Real-Time Monitoring communication quality;

(3), invention increases the design of built-in function; In signal process function storehouse, prestore multi-signal process function, user can realize processing in real time extracting the bioelectrical signals obtained under all kinds of environment without the need to programming; Secondly, signal process function storehouse can be expanded, and can write the signal process function that user writes voluntarily, meets all kinds of demands of user;

(4), in signal processing interface, user can as required, and in selectivity call signal process function library, function processes bioelectrical signals, and operating process is simple.

Accompanying drawing explanation

Fig. 1 is a kind of detailed description of the invention Organization Chart of bioelectrical signals monitoring system that the present invention is based on LabVIEW;

Fig. 2 is a kind of detailed description of the invention structure chart of the extraction equipment of bioelectrical signals shown in Fig. 1;

Fig. 3 is a kind of detailed description of the invention structure chart of the observation interface of bioelectrical signals shown in Fig. 1.

Detailed description of the invention

Below in conjunction with accompanying drawing, the specific embodiment of the present invention is described, so that those skilled in the art understands the present invention better.Requiring particular attention is that, in the following description, when perhaps the detailed description of known function and design can desalinate main contents of the present invention, these are described in and will be left in the basket here.

Embodiment

LabVIEW (Laboratory Virtual Instrumentation Engineering Workbench): laboratory virtual instrument engineering platform;

Zigbee: based on a low-power consumption territory fidonetFido of IEEE802.15.4 standard;

Fig. 1 is a kind of detailed description of the invention Organization Chart of bioelectrical signals monitoring system that the present invention is based on LabVIEW.

In the present embodiment, as shown in Figure 1, a kind of bioelectrical signals monitoring system based on LabVIEW of the present invention, comprising: bioelectrical signals extraction equipment 1, electromagnetic shield signal line 2, PCB circuit integrated module 3, zigbee group-net communication module 4, signal process function storehouse 5 and bioelectrical signals observation interface 6.

As shown in Figure 2, bioelectrical signals extraction equipment 1 comprises again electroencephalogramsignal signal collection equipment and multiple limbs signal collecting device, and wherein, Fig. 2 left side is electroencephalogramsignal signal collection equipment, is limbs signal collecting device on the right of Fig. 2.Spring steel plate 1.1, stretchy nylon band 1.2 and dry electrode slice 1.3 is included in two collecting devices; Wherein, in electroencephalogramsignal signal collection equipment, dry electrode slice 1.3 has 3 pieces at least.Dry electrode slice 1.3, can be stuck in well on head and limbs, and welding position is corresponding with bioelectrical signals active regions by solder attachment on the spring steel plate 1.1 of correspondence; Stretchy nylon band 1.2 can add capable and experienced electrode slice 1.3 and the closeness of contact of human body further, increases the stability of acquiring biological electric signals; When extracting bioelectrical signals, electroencephalogramsignal signal collection equipment being worn on head, multiple limbs signal collecting device being worn on respectively wrist, the large position such as arm and little arm, extracting the bioelectrical signals on brain and limbs respectively; All be connected with an electromagnetic shield signal line 2 at each collecting device end, the bioelectrical signals of collection sends to PCB circuit integrated module by electromagnetic shield signal line 2 again;

Electromagnetic shield signal line 2, adopt the anti-tampering design of multilamellar, wherein internal layer is signal transmssion line, skin is rubber layer, is tinsel envelope, thus realizes electromagnetic shielding to a certain extent in the periphery of signal transmssion line, enhance the capacity of resisting disturbance of the signal of telecommunication, increase effective propagation path;

PCB circuit integrated module 3, comprises electromagnetic interface filter, preamplifier and analog-digital converter; For receiving the multichannel bioelectrical signals that bioelectrical signals extraction equipment 1 gathers, and by the bioelectrical signals parallel processing process of multichannel, then send to zigbee group-net communication module 4;

After PCB circuit integrated module 3 receives bioelectrical signals, first pass through the electromagnetic noise that the outer frequency of electromagnetic interface filter filter out-band is higher, filtered bioelectrical signals is input to analog-digital converter after preamplifier amplifies, the bioelectrical signals of simulation is converted into the digital signal being easy to receive and dispatch by analog-digital converter again, and sends to zigbee group-net communication module 4;

Zigbee group-net communication module contains multiple port, can walk abreast and receive the multichannel bioelectrical signals of PCB circuit integrated module 3 transmission, then be transmitted to bioelectrical signals observation interface 6, realizes the network type communication that many receipts are multiple;

Signal process function storehouse 5, signal processing interface, by function in call signal process function library, processes bioelectrical signals, re-sends to bioelectrical signals observation interface 6; In the present embodiment, signal process function storehouse includes the many kinds of functions such as filter function, trap function, wavelet function and artificial neural network function;

As shown in Figure 3, bioelectrical signals observation interface 6 comprises again signal monitoring interface and signal processing interface; By the tab control 6.1 on bioelectrical signals observation interface, select entering signal observation interface or signal processing interface, wherein, Fig. 3 (a) is signal monitoring interface, and Fig. 3 (b) is signal processing interface;

On signal monitoring interface, at 6.2 places, the COM1 between zigbee group-net communication module 4 and bioelectrical signals observation interface 6 can be set, select signalling channel when needing entering signal process interface, when communicating, the signal quality of bioelectrical signals directly can be observed on instrumental panel 6.3, the relevant information of bioelectrical signals demonstrates the bioelectrical signals on brain and limbs respectively by 2-D display box, namely 6.5 and 6.6 places demonstrate the bioelectrical signals on brain and limbs respectively;

In the present embodiment, as shown in Fig. 3 (a), instrumental panel 6.3 point 0---10 scales, when the pointer on instrumental panel 6.3 refers to certain scale, represent the degree of this bioelectrical signals by external environmental interference, wherein, " 0 " is desirable interference-free communication, and " 10 " table is that external environmental interference is violent;

On signal processing interface, by pending signalling channels such as display lamp 6.10 show, user can be as required, function in selectivity call signal process function library, namely can call wherein some or multiple function to process the bioelectrical signals in this signalling channel, wherein, when specifically calling certain function, arrange corresponding function, concrete setting procedure is as follows again:

1), at 6.7 places can call the filter function in signal process function storehouse, and set out and leach frequency range

In the present embodiment, wave filter in built-in function is butterworth filter (Butterworth), user can arrange filtering frequency range as required, extracts the signal specific frequency range wanting deep monitoring, or for filtering low frequency spur and high frequency supurious wave; Such as, think to observe separately in EEG signals alpha frequency range (8Hz-12Hz) brain wave embodying Mental imagery, then can, on the basis of original EEG signals, use filter function to leach the EEG signals of this frequency range;

2), at 6.8 places can call the trap function in signal process function storehouse, set out the Hz noise frequency range that trap function is removed

In the present embodiment, realize trap by the band elimination filter based on FIR, trap frequency is 50Hz or 60Hz;

3), at 6.9 places can call the wavelet function in signal process function storehouse, the female ripple type of setting small echo, selects the number of plies of wavelet decomposition

In the present embodiment, mallat algorithm is adopted to build wavelet function, its type is the female ripple of haar or db small echo, and by the setting wavelet decomposition number of plies, to the wavelet Smoothing process that primary signal is carried out in various degree, Decomposition order is more in principle, smoothing processing effect is better, but likely can the useful characteristic information of lost part, at this, decomposition level selects 3 layers;

4) the neural network classification function, in all right call signal process function library of system, the parameter of neural network classifier is without the need to arranging, need train and obtain, because different human body bioelectrical signals reference energy difference is larger, therefore before calling neural network classification function, need train neural network classifier parameter, training step is as follows:

S1: when kth training study, extracts the large number of biological signal of telecommunication as training sample data from monitored person, is labeled as: x (k)=(x ₁(k), x ₂(k) ..., x _n(k)), wherein, k=1,2 ..., m, m ∈ M;

S2: during serial communication under 38400 baud rates, training sample data are sent on signal processing interface; Now signal processing interface is defaulted as Neural Networks Training Pattern;

S3: the parameter training completing neural network classifier

Training sample data are read at signal processing interface, according to sample data, neural network classifier modeling parameters is sent under the serial communication of 38400 baud rates, neural network model parameter is set, comprise input layer number n, hidden layer neuron number p, output layer neuron number q, allowable error ε and maximum study number of times M;

Obtained by the computed in software in step S2:

Hidden layer inputs: ${\mathrm{hi}}_h\left(k\right)=\underset{i=1}{\overset{n}{\mathrm{\Σ}}}{w}_{\mathrm{ih}}{x}_i\left(k\right)-b_h,h=\mathrm{1,2},...,p;$

Hidden layer exports: ho _h(k)=f (hi _h(k)) h=1,2 ..., p;

Output layer inputs: ${\mathrm{yi}}_o\left(k\right)=\underset{h=1}{\overset{p}{\mathrm{\Σ}}}{w}_{\mathrm{ho}}{\mathrm{ho}}_h\left(k\right)-b_o,o=\mathrm{1,2},...q;$

Output layer exports: yo _o(k)=f (yi _o(k)) o=1,2 ... q;

Error function: $e=\frac{1}{2}\underset{o=1}{\overset{q}{\mathrm{\Σ}}}{(d_o\left(k\right)-{\mathrm{yo}}_o\left(k\right))}^2;$

Wherein, b _hand b _ofor constant vector, vector length is identical with hidden layer/output layer neuron number; w _hothat hidden layer connects weights; w _ihinput layer connects weights; F () is the S type function that can lead, net=x ₁w ₁+ x ₂w ₂+ ...+x _nw _n, wherein, x _n, w _nthe input being respectively f () be connected weights;

If desired output: d _o(k)=(d ₁(k), d ₂(k) ..., d _q(k)), export and desired output according to output layer, try to achieve error function, then calculate the partial derivative of error function to output layer neuron and hidden layer neuron, that is:

The local derviation that error function inputs output layer:

$\begin{array}{c}\frac{\∂e}{{\∂\mathrm{yi}}_o}=\frac{\∂{\left(\frac{1}{2}\underset{o=1}{\overset{q}{\mathrm{\Σ}}}(d_o\left(k\right)-{\mathrm{yo}}_o\left(k\right))\right)}^2}{{\∂\mathrm{yi}}_o}=-(d_o\left(k\right)-{\mathrm{yo}}_o\left(k\right)){\mathrm{yo}}_o^{\′}\left(k\right)\\ =-(d_o\left(k\right)-{\mathrm{yo}}_o\left(k\right))f^{\′}\left({\mathrm{yi}}_o\left(k\right)\right)\stackrel{\mathrm{\Δ}}{=}-{\mathrm{\δ}}_o\left(k\right)\end{array}$

Wherein, for substitute symbol, namely use-δ _ok () substitutes-(d _o(k)-yo _o(k)) f ' (yi _o(k));

Output layer input connects the local derviation of weights to hidden layer:

$\frac{{\∂\mathrm{yi}}_o\left(k\right)}{{\∂w}_{\mathrm{ho}}}=\frac{\∂(\underset{h}{\overset{p}{\mathrm{\Σ}}}{w}_{\mathrm{ho}}{\mathrm{ho}}_h\left(k\right)-b_o)}{{\∂w}_{\mathrm{ho}}}={\mathrm{ho}}_h\left(k\right)$

Error function connects the local derviation of weights to hidden layer:

$\frac{\∂e}{{\∂w}_{\mathrm{ho}}}=\frac{\∂e}{{\∂\mathrm{yi}}_o}\frac{{\∂\mathrm{yi}}_o}{{\∂w}_{\mathrm{ho}}}=-{\mathrm{\δ}}_o\left(k\right){\mathrm{ho}}_h\left(k\right)$

The local derviation that error function inputs hidden layer:

$\begin{array}{c}\frac{\∂e}{{\∂\mathrm{hi}}_h\left(k\right)}=\frac{\∂\left(\frac{1}{2}\underset{o=1}{\overset{q}{\mathrm{\Σ}}}{(d_o\left(k\right)-{\mathrm{yo}}_o\left(k\right))}^2\right)}{{\∂\mathrm{ho}}_h\left(k\right)}\frac{{\∂\mathrm{ho}}_h\left(k\right)}{{\∂\mathrm{hi}}_h\left(k\right)}\\ =\frac{\∂\left(\frac{1}{2}\underset{o=1}{\overset{q}{\mathrm{\Σ}}}{(d_o\left(k\right)-f\left({\mathrm{yi}}_o\left(k\right)\right))}^2\right)}{{\∂\mathrm{ho}}_h\left(k\right)}\frac{{\∂\mathrm{ho}}_h\left(k\right)}{{\∂\mathrm{hi}}_h\left(k\right)}=\frac{\∂\left(\frac{1}{2}\underset{o=1}{\overset{q}{\mathrm{\Σ}}}\left(\right(d_o\left(k\right)-f{(\underset{h=1}{\overset{p}{\mathrm{\Σ}}}{w}_{\mathrm{ho}}{\mathrm{ho}}_h\left(k\right)-b_o)}^2)\right)}{{\∂\mathrm{ho}}_h\left(k\right)}\frac{{\∂\mathrm{ho}}_h\left(k\right)}{{\∂\mathrm{hi}}_h\left(k\right)}\\ =-\underset{o=1}{\overset{q}{\mathrm{\Σ}}}(d_o\left(k\right)-{\mathrm{yo}}_o\left(k\right))f^{\′}\left({\mathrm{yi}}_o\left(k\right)\right)w_{\mathrm{ho}}\frac{{\∂\mathrm{ho}}_h\left(k\right)}{{\∂\mathrm{hi}}_h\left(k\right)}=-\left(\underset{o=1}{\overset{q}{\mathrm{\Σ}}}{\mathrm{\δ}}_o\left(k\right)w_{\mathrm{ho}}\right)f^{\′}\left({\mathrm{hi}}_h\left(k\right)\right)\stackrel{\mathrm{\Δ}}{=}-{\mathrm{\δ}}_h\left(k\right)\end{array}$

Error function connects the local derviation of weights to input layer:

$\frac{\∂e}{{\∂w}_{\mathrm{ih}}}=\frac{\∂e}{{\∂\mathrm{hi}}_h\left(k\right)}\frac{{\∂\mathrm{hi}}_h\left(k\right)}{{\∂w}_{\mathrm{ih}}}$

Hidden layer input connects the local derviation of weights to input layer:

$\frac{{\∂\mathrm{hi}}_h\left(k\right)}{{\∂w}_{\mathrm{ih}}}=\frac{\∂(\underset{i=1}{\overset{n}{\mathrm{\Σ}}}{w}_{\mathrm{ih}}{x}_i\left(k\right)-b_h)}{{\∂w}_{\mathrm{ih}}}=x_i\left(k\right)$

According to above-mentioned local derviation equation, revise the connection weights that hidden layer connects weights and input layer

Hidden layer connects weights: $\begin{array}{c}{\mathrm{\Δw}}_{\mathrm{ho}}\left(k\right)=-\mathrm{\μ}\frac{\∂e}{{\∂w}_{\mathrm{ho}}}=\mathrm{\μ}{\mathrm{\δ}}_o\left(k\right){\mathrm{ho}}_h\left(k\right)\\ w_{\mathrm{ho}}^{k+1}=w_{\mathrm{ho}}^k+\mathrm{\η}{\mathrm{\δ}}_o\left(k\right){\mathrm{ho}}_h\left(k\right)\end{array}$

Input layer connects weights: $\begin{array}{c}{\mathrm{\Δw}}_{\mathrm{ih}}\left(k\right)=-\mathrm{\μ}\frac{\∂e}{{\∂w}_{\mathrm{ih}}}=-\mathrm{\μ}\frac{\∂e}{{\∂\mathrm{hi}}_h\left(k\right)}\frac{{\∂\mathrm{hi}}_h\left(k\right)}{{\∂w}_{\mathrm{ih}}}={\∂}_h\left(k\right)x_i\left(k\right)\\ w_{\mathrm{ih}}^{k+1}=w_{\mathrm{ih}}^k+\mathrm{\η}{\mathrm{\δ}}_h\left(k\right)x_i\left(k\right)\end{array}$

μ, η are constant, are all set to 0.01 in the present embodiment; the hidden layer obtained during training study secondary to kth connects weights, the input layer obtained during training study secondary to kth connects weights; To now, the number of times m having completed training study will increase once;

Calculate global error during front k training study, global error:

Judge whether global error reaches default precision or study number of times m reaches upper limit M, if reached, then neural network classifier parameter training terminates; Otherwise return step S1, carry out kth+1 training study.

Wherein, neural network classifier cannot use different people, even cannot use when different conditions same person, so after signal processing interface is exited completely, to automatically remove this training result, and in time using next time, then need Resurvey training data to train;

Bioelectrical signals is after above-mentioned process completes, and its relevant information demonstrates the bioelectrical signals on brain and limbs respectively by 2-D display box, and namely 6.11 and 6.12 places demonstrate the bioelectrical signals on brain and limbs respectively.

Although be described the illustrative detailed description of the invention of the present invention above; so that those skilled in the art understand the present invention; but should be clear; the invention is not restricted to the scope of detailed description of the invention; to those skilled in the art; as long as various change to limit and in the spirit and scope of the present invention determined, these changes are apparent, and all innovation and creation utilizing the present invention to conceive are all at the row of protection in appended claim.

Claims (5)

1., based on a bioelectrical signals monitoring system of LabVIEW, it is characterized in that, comprising:

One bioelectrical signals extraction equipment, comprise electroencephalogramsignal signal collection equipment and multiple limbs signal collecting device, in two collecting devices, include spring steel plate, stretchy nylon band and dry electrode slice, wherein, in electroencephalogramsignal signal collection equipment, at least comprise 3 dry electrode slices; Dry electrode slice, is stuck in during by wearing bioelectrical signals extraction equipment on head and limbs by solder attachment on the spring steel plate of correspondence, and welding position is corresponding with bioelectrical signals active regions; Stretchy nylon band is positioned on spring steel plate, can add the closeness of contact of capable and experienced electrode slice and human body further, increases the stability of acquiring biological electric signals; When extracting bioelectrical signals, bioelectrical signals extraction equipment is worn on head and each position of limbs, extract the bioelectrical signals on brain and each position of limbs respectively, then send to PCB circuit integrated module by the electromagnetic shield signal line that collecting device connects;

Many electromagnetic shield signal lines, adopt the anti-tampering design of multilamellar, wherein internal layer is signal transmssion line, and skin is rubber layer, is tinsel envelope in the periphery of signal transmssion line;

One PCB circuit integrated module, comprises electromagnetic interface filter, preamplifier and analog-digital converter; For receiving the multichannel bioelectrical signals that bioelectrical signals extraction equipment gathers, and by the bioelectrical signals parallel processing of multichannel, then send to zigbee group-net communication module;

After PCB circuit integrated module receives bioelectrical signals, first pass through the electromagnetic noise that the outer frequency of electromagnetic interface filter filter out-band is higher, filtered bioelectrical signals is input to analog-digital converter after preamplifier amplifies, the bioelectrical signals of simulation is converted into the digital signal being easy to receive and dispatch by analog-digital converter again, and sends to zigbee group-net communication module;

One zigbee group-net communication module contains multiple COM1, can walk abreast and receive the multichannel bioelectrical signals of PCB circuit integrated module transmission, then be transmitted to bioelectrical signals observation interface, realizes the network type communication that many receipts are multiple;

One signal process function storehouse, signal processing interface, by function in call signal process function library, processes bioelectrical signals, re-sends to bioelectrical signals observation interface;

One bioelectrical signals observation interface, by the tab control on bioelectrical signals observation interface, selects entering signal observation interface or signal processing interface;

On signal monitoring interface, COM1 between zigbee group-net communication module and bioelectrical signals observation interface can be set, select signalling channel when needing entering signal process interface, when communicating, the signal quality of bioelectrical signals directly can be observed on instrumental panel, and the relevant information of bioelectrical signals demonstrates the bioelectrical signals on brain and limbs respectively by 2-D display box;

On signal processing interface, user can function as required in selectivity call signal process function library, then arranges according to the function called, and as Choose filtering function, then sets out and leaches frequency range; As selected trap function, then set out the Hz noise frequency range that trap function is removed; As selected wavelet function, then set the female ripple type of small echo, select the number of plies of wavelet decomposition; As selected neural network function, then when calling this function, train neural network classifier parameter; Bioelectrical signals is after above-mentioned process completes, and its relevant information demonstrates the bioelectrical signals on brain and limbs respectively by 2-D display box.

2. the bioelectrical signals monitoring system based on LabVIEW according to claim 1, is characterized in that, at least comprises in described signal process function storehouse: filter function, trap function, wavelet function and artificial neural network function.

3. the bioelectrical signals monitoring system based on LabVIEW according to claim 1, it is characterized in that, described instrumental panel divides 0---10 scales, when the pointer on instrumental panel refers to certain scale, represent the degree of this bioelectrical signals by external environmental interference, wherein, " 0 " is desirable interference-free communication, and " 10 " table is that external environmental interference is violent.

4. the bioelectrical signals monitoring system based on LabVIEW according to claim 1, is characterized in that, described neural network training classifier parameters method is:

(4.1), when kth training study, extract the large number of biological signal of telecommunication as training sample data from monitored person, be labeled as: x (k)=(x ₁(k), x ₂(k) ..., x _n(k)), wherein, k=1,2 ..., m, m ∈ M;

(4.2), during serial communication, under 38400 baud rates, training sample data are sent on signal processing interface;

(4.3) parameter training of neural network classifier, is completed

Training sample data are read at signal processing interface, according to sample data, neural network classifier modeling parameters is sent under the serial communication of 38400 baud rates, neural network model parameter is set, comprise input layer number n, hidden layer neuron number p, output layer neuron number q, allowable error ε and maximum study number of times M;

By calculating:

Hidden layer inputs: ${\mathrm{hi}}_h\left(k\right)=\underset{i=1}{\overset{n}{\mathrm{\Σ}}}{w}_{\mathrm{ih}}{x}_i\left(k\right)-b_h$ h＝1,2,…,p；

Hidden layer exports: ho _h(k)=f (hi _h(k)) h=1,2 ..., p;

Output layer inputs: ${\mathrm{yi}}_o\left(k\right)=\underset{h=1}{\overset{p}{\mathrm{\Σ}}}{w}_{\mathrm{ho}}{\mathrm{ho}}_h\left(k\right)-b_o$ o＝1,2,…q；

Output layer exports: yo _o(k)=f (yi _o(k)) o=1,2 ... q;

Error function: $e=\frac{1}{2}\underset{o=1}{\overset{q}{\mathrm{\Σ}}}{(d_o\left(k\right)-{\mathrm{yo}}_o\left(k\right))}^2;$

Wherein, f () is the S type function that can lead;

If desired output: d _o(k)=(d ₁(k), d ₂(k) ..., d _q(k)), export and desired output according to output layer, try to achieve error function, then calculate error function to b _hand b _ofor constant vector, vector length is identical with hidden layer/output layer neuron number; w _hothat hidden layer connects weights; w _ihinput layer connects weights; The partial derivative of output layer neuron and hidden layer neuron, that is:

The local derviation that error function inputs output layer:

$\begin{array}{c}\frac{\∂e}{\∂{\mathrm{yi}}_o}=\frac{\∂{\left(\frac{1}{2}\underset{o=1}{\overset{q}{\mathrm{\Σ}}}(d_o\left(k\right)-{\mathrm{yo}}_o\left(k\right))\right)}^2}{\∂{\mathrm{yi}}_o}=-(d_o\left(k\right)-{\mathrm{yo}}_o\left(k\right)){\mathrm{yo}}_o^{\′}\left(k\right)\\ =-(d_o\left(k\right)-{\mathrm{yo}}_o\left(k\right))f^{\′}\left({\mathrm{yi}}_o\left(k\right)\right)\stackrel{\mathrm{\Δ}}{=}-{\mathrm{\δ}}_o\left(k\right)\end{array}$

Output layer input connects the local derviation of weights to hidden layer:

$\frac{\∂{\mathrm{yi}}_o\left(k\right)}{\∂w_{\mathrm{ho}}}=\frac{\∂(\underset{h=1}{\overset{p}{\mathrm{\Σ}}}{w}_{\mathrm{ho}}{\mathrm{ho}}_h\left(k\right)-b_o)}{\∂w_{\mathrm{ho}}}={\mathrm{ho}}_h\left(k\right)$

Error function connects the local derviation of weights to hidden layer:

$\frac{\∂e}{\∂w_{\mathrm{ho}}}=\frac{\∂e}{\∂{\mathrm{yi}}_o}\frac{\∂{\mathrm{yi}}_o}{\∂w_{\mathrm{ho}}}=-{\mathrm{\δ}}_o\left(k\right){\mathrm{ho}}_h\left(k\right)$

The local derviation that error function inputs hidden layer:

$\begin{array}{c}\frac{\∂e}{\∂{\mathrm{hi}}_h\left(k\right)}=\frac{\∂{\left(\frac{1}{2}\underset{o=1}{\overset{q}{\mathrm{\Σ}}}(d_o\left(k\right)-{\mathrm{yo}}_o\left(k\right))\right)}^2}{\∂{\mathrm{ho}}_h\left(k\right)}\frac{\∂{\mathrm{ho}}_h\left(k\right)}{\∂{\mathrm{hi}}_h\left(k\right)}\\ ==-\left(\underset{o=1}{\overset{q}{\mathrm{\Σ}}}{\mathrm{\δ}}_o\left(k\right)w_{\mathrm{ho}}\right)f^{\′}\left({\mathrm{hi}}_h\left(k\right)\right)\stackrel{\mathrm{\Δ}}{=}-{\mathrm{\δ}}_h\left(k\right)\end{array}$

for substitute symbol;

Error function connects the local derviation of weights to input layer:

$\frac{\∂e}{\∂w_{\mathrm{ih}}}=\frac{\∂e}{\∂{\mathrm{hi}}_h\left(k\right)}\frac{\∂{\mathrm{hi}}_h\left(k\right)}{\∂w_{\mathrm{ih}}}$

Hidden layer input connects the local derviation of weights to input layer:

$\frac{\∂{\mathrm{hi}}_h\left(k\right)}{\∂w_{\mathrm{ih}}}=\frac{\∂(\underset{i=1}{\overset{n}{\mathrm{\Σ}}}{w}_{\mathrm{ih}}{x}_i\left(k\right)-b_h)}{\∂w_{\mathrm{ih}}}=x_i\left(k\right)$

According to above-mentioned local derviation equation, revise the connection weights that hidden layer connects weights and input layer

Hidden layer connects weights: $\begin{array}{c}\mathrm{\Δ}{w}_{\mathrm{ho}}\left(k\right)=-\mathrm{\μ}\frac{\∂e}{\∂w_{\mathrm{ho}}}=\mathrm{\μ}{\mathrm{\δ}}_o\left(k\right){\mathrm{ho}}_h\left(k\right)\\ w_{\mathrm{ho}}^{k+1}=w_{\mathrm{ho}}^k+\mathrm{\η}{\mathrm{\δ}}_o\left(k\right){\mathrm{ho}}_h\left(k\right)\end{array}$

Input layer connects weights: $\begin{array}{c}\mathrm{\Δ}{w}_{\mathrm{ih}}\left(k\right)=-\mathrm{\μ}\frac{\∂e}{\∂w_{\mathrm{ih}}}=-\mathrm{\μ}\frac{\∂e}{\∂{\mathrm{hi}}_h\left(k\right)}\frac{\∂{\mathrm{hi}}_h\left(k\right)}{\∂w_{\mathrm{ih}}}={\mathrm{\δ}}_h\left(k\right)x_i\left(k\right)\\ w_{\mathrm{ih}}^{k+1}=w_{\mathrm{ih}}^k+\mathrm{\η}{\mathrm{\δ}}_h\left(k\right)x_i\left(k\right)\end{array}$

μ, η are constant, the hidden layer obtained during training study secondary to kth connects weights, the input layer obtained during training study secondary to kth connects weights;

Calculate global error during front k training study, global error:

Judge whether global error reaches default precision or study number of times m reaches upper limit M, if reached, then neural network classifier parameter training terminates; Otherwise return step (4.1), carry out kth+1 training study.

5. the bioelectrical signals monitoring system based on LabVIEW according to claim 1 or 3, it is characterized in that, described neural network classifier, after signal processing interface is exited completely, will remove this training result automatically, in time using next time, then Resurvey training data is needed to train.