CN104133404A - Method and device for processing signal - Google Patents

Abstract

The invention provides a method for processing a signal. The method includes the steps of conditioning the original signal according to preset conditioning conditions to obtain an analog signal; carrying out A/D conversion on the analog signal to obtain a digital signal; carrying out spectral analysis on the digital signal with an FIR algorithm and an FFT algorithm to obtain a first signal; correcting the frequency and the amplitude of the first signal with a preset correcting algorithm to obtain a second signal; compensating for the amplitude of the second signal according to a preset compensation rule to obtain a third signal, and outputting the third signal to a calculation device. By means of the method, after spectral analysis is carried out on the digital signal with the FIR algorithm and the FFT algorithm, the frequency and the amplitude are corrected, and the amplitude is compensated. The signal is corrected, spectrum leakage of the FFT algorithm is reduced, and meanwhile the picket fence effect of the FFT algorithm is buffered; amplitude compensation is carried out on the signal, FIR band-pass fluctuation and transition band attenuation are overcome, and the accuracy of the signal processing result is improved.

Description

A kind of signal processing method and device

Technical field

The present invention relates to signal process field, in particular, relate to a kind of signal processing method and device.

Background technology

Be accompanied by the development of large scale integrated circuit and the further investigation of control algolithm, digital signal processing is high with its precision, reliability is high, dirigibility is large, be easy to the advantages such as large-scale integrated, in every field, is all used widely.

Digital signal processing is to adopt the method for numerical evaluation to process Serial No., signal is transformed into certain form suiting the requirements.As, digital signal is carried out to filtering to limit other frequency bands or filtering noise and interference, or they are carried out separatedly with other signals, signal is carried out to spectrum analysis and to understand the frequency spectrum of signal, form, and then signal is identified.

In prior art, the main method that FIR (Finite impact Response, has limit for length's impulse response) algorithm and FFT (Fast Fourier Transform, fast fourier transform) algorithm are spectrum analysis.Traditional signal processing flow as shown in Figure 1, original signal is delivered to digital signal processing unit and is carried out spectrum analysis after changing via signal condition, A/D (Analog-to-digital, analog/digital), then calculates.Wherein, this spectrum analysis link is comprised of FIR and FFT.

Because low order FIR exists the logical fluctuation of band greatly and the problem of transitional zone decay, cause the signal attenuation coefficient of different frequency point different, can cause the signal erroneous judgement of critical frequency.And there is the spectrum leakage problem that affects measurement accuracy in fft algorithm when application.

Therefore, adopt method of the prior art when carrying out signal processing, easily occur the problem that signal processing results precision is lower.

Summary of the invention

In view of this, the invention provides a kind of signal processing method and device, solved the lower problem of signal processing results precision in prior art.

For achieving the above object, the invention provides following technical scheme:

A signal processing method, comprising:

According to default conditioning condition, original signal is nursed one's health, obtained simulating signal;

Described simulating signal is carried out to analog/digital A/D conversion, obtain digital signal;

According to having limit for length's impulse response FIR algorithm and fast fourier transform fft algorithm to carry out spectrum analysis to described digital signal, obtain first signal;

According to default correction algorithm, the frequency of described first signal and amplitude are carried out to correcting process, obtain secondary signal;

According to default offset rule, the amplitude compensation to described secondary signal, obtains the 3rd signal and exports computing equipment to.

Above-mentioned method, preferred, described correction algorithm comprises FFT difference correction algorithm, the default correction algorithm of described foundation carries out correcting process to the frequency of described first signal and amplitude and specifically comprises:

According to default FFT interpolation algorithm, calculate the discrete spectrum of described secondary signal;

Select the spectral line of default number in described discrete spectrum;

Calculate the spectral line amplitude that each spectral line is corresponding;

The window function that foundation is default and described spectral line amplitude and default fitting condition calculate update information, and according to described update information, the frequency of described first signal and amplitude are revised, and obtain secondary signal.

Above-mentioned method, preferred, described foundation is preset offset rule, and the amplitude compensation of described secondary signal is specifically comprised:

From default FIR compensating coefficient table, search the compensated information corresponding with the frequency of described secondary signal;

According to described compensated information, to the compensation of tabling look-up of the amplitude of described secondary signal, obtain the 3rd signal.

Above-mentioned method, preferred, in the described discrete spectrum of described selection, the spectral line of default number comprises:

Determine the spectral line of amplitude maximum in described discrete spectral line;

Select remaining spectral line in the described default number adjacent with described maximum spectral line.

Above-mentioned method, preferred, the window function that described foundation is default and described spectral line amplitude, and default fitting condition, comprise the frequency of described first signal and amplitude correction:

According to described amplitude, construct respectively the first function and the second function;

Fitting function according to described default window function, MATLAB, respectively to described the first function and the second Function Fitting, obtain the actual frequency value function of described first signal and the fitting polynomial formulas of actual magnitude function, and utilize described fitting polynomial formulas to carry out correcting process to the frequency of described first signal and amplitude.

Above-mentioned method, preferred, described window function adopts Blackman window function.

Above-mentioned method, preferred, in described definite described discrete spectral line, the spectral line of amplitude maximum comprises:

With described predeterminated position, be as the criterion, described discrete spectral line is adopted to the sequence of bubble sort method, obtain the spectral line of amplitude maximum.

Above-mentioned method, preferred, described default number comprises: 2,3 or 5.

A signal processing apparatus, comprising:

Conditioning module, for according to default conditioning condition, original signal being nursed one's health, obtains simulating signal;

Modular converter, for described simulating signal is carried out to A/D conversion, obtains digital signal;

Spectrum analysis module, for according to FIR algorithm and fft algorithm, described digital signal being carried out to spectrum analysis, obtains first signal;

Correcting module, for according to default correction algorithm, the frequency of described first signal and amplitude being carried out to correcting process, obtains secondary signal;

Compensating module, for the default offset rule of foundation, the amplitude compensation to described secondary signal, obtains the 3rd signal and exports computing equipment to.

Above-mentioned device, preferred, described correction algorithm comprises FFT difference correction algorithm, described correcting module specifically for:

According to default FFT interpolation algorithm, calculate the discrete spectrum of described secondary signal;

Select the spectral line of default number in described discrete spectrum;

Calculate the spectral line amplitude that each spectral line is corresponding;

The window function that foundation is default and described spectral line amplitude and default fitting condition calculate update information, and according to described update information, the frequency of described first signal and amplitude are revised, and obtain secondary signal.

Known via above-mentioned technical scheme, compared with prior art, the invention provides a kind of signal processing method, comprising: according to default conditioning condition, original signal is nursed one's health, obtained simulating signal; Described simulating signal is carried out to A/D conversion, obtain digital signal; According to FIR algorithm and fft algorithm, described digital signal is carried out to spectrum analysis, obtain first signal; According to default correction algorithm, the frequency of described first signal and amplitude are carried out to correcting process, obtain secondary signal; According to default offset rule, the amplitude compensation to described secondary signal, obtains the 3rd signal and exports computing equipment to.Adopt the method, after digital signal being carried out to spectrum analysis according to FIR algorithm and fft algorithm and completing, proceed the correcting process of frequency and amplitude, and amplitude is compensated.Signal is carried out to correcting process, subdued the spectrum leakage phenomenon of fft algorithm, alleviated the fence effect of fft algorithm simultaneously; And signal is carried out to amplitude compensation, and made up the problem of the logical fluctuation of FIR band and transitional zone decay, improved the precision of signal processing results.

Accompanying drawing explanation

In order to be illustrated more clearly in the embodiment of the present invention or technical scheme of the prior art, to the accompanying drawing of required use in embodiment or description of the Prior Art be briefly described below, apparently, accompanying drawing in the following describes is only embodiments of the invention, for those of ordinary skills, do not paying under the prerequisite of creative work, other accompanying drawing can also be provided according to the accompanying drawing providing.

Fig. 1 is traditional signal processing flow figure;

The process flow diagram of a kind of signal processing method embodiment 1 that Fig. 2 provides for the application;

The process flow diagram of a kind of signal processing method embodiment 2 that Fig. 3 provides for the application;

The particular flow sheet of step S307 in a kind of signal processing method embodiment 2 that Fig. 4 provides for the application;

The process flow diagram of a kind of signal processing method embodiment 3 that Fig. 5 provides for the application;

The response diagram of example in a kind of signal processing method embodiment 3 that Fig. 6 provides for the application;

The frequency of example and the graph of a relation between amplitude attenuation coefficient in a kind of signal processing method embodiment 3 that Fig. 7 provides for the application;

The structural representation of a kind of signal processing apparatus embodiment 1 that Fig. 8 provides for the application.

Embodiment

Below in conjunction with the accompanying drawing in the embodiment of the present invention, the technical scheme in the embodiment of the present invention is clearly and completely described, obviously, described embodiment is only the present invention's part embodiment, rather than whole embodiment.Embodiment based in the present invention, those of ordinary skills, not making the every other embodiment obtaining under creative work prerequisite, belong to the scope of protection of the invention.

The process flow diagram of a kind of signal processing method that the embodiment of the present application 1 as shown in Figure 2 provides, comprising:

Step S201: according to default conditioning condition, original signal is nursed one's health, obtained simulating signal;

Wherein, this conditioning condition can be the condition of follow-up A/D transformational structure input signal.

Concrete, this conditioning process comprises: power amplification, filtering, linear compensation, isolation and/or protection etc., and the process of conditioning can be set according to actual conditions, and the present embodiment does not repeat.

Step S202: described simulating signal is carried out to A/D conversion, obtain digital signal;

Wherein, because follow-up FIR algorithm and fft algorithm need to be according to digital format signals, and what after conditioning, obtain is analog format signal, therefore, this simulating signal is carried out to A/D conversion, makes this signal be converted to simulating signal.

Step S203: according to FIR algorithm and fft algorithm, described digital signal is carried out to spectrum analysis, obtain first signal;

Wherein, according to FIR algorithm and fft algorithm, successively digital signal is carried out to spectrum analysis, obtain first signal.

Concrete, this spectrum analysis process is: adopt fft algorithm to realize the conversion to digital signal, adopt FIR algorithm to realize the filtering to digital signal.

But, adopt fft algorithm, in this conversion process, easily there is spectrum leakage phenomenon; And there is the logical problem large and transitional zone decay that fluctuates of band in the filtering that adopts FIR algorithm.

Therefore, at follow-up step S204-205, so that the problems referred to above are made up.

Step S204: according to default correction algorithm, the frequency of described first signal and amplitude are carried out to correcting process, obtain secondary signal;

Wherein, before carrying out signal processing, default correction algorithm.

Concrete, according to this correction algorithm, the frequency of first signal and amplitude are revised respectively the problem of the spectrum leakage of appearance while carrying out digital conversion to overcome fft algorithm.

Concrete makeover process describes in follow-up embodiment, and the present embodiment does not repeat.

Step S205: according to default offset rule, the amplitude compensation to described secondary signal, obtains the 3rd signal and export computing equipment to.

Wherein, before carrying out signal processing, default offset rule.

Concrete, according to this offset rule, to carrying out amplitude compensation through the secondary signal of correction, to overcome the problem that the logical fluctuation of the band that adopts the filtering of FIR algorithm to cause is large and transitional zone decays.

Wherein, the 3rd signal is exported to follow-up computing equipment, to carry out follow-up computation process.

It should be noted that, while digital signal being carried out to spectrum analysis in the application, what adopt is FIR algorithm and fft algorithm, in actual enforcement, when carrying out spectrum analysis, also can only adopt a kind of in FIR algorithm and fft algorithm, accordingly, when adopt be above-mentioned two kinds of algorithms time, perform step S204 and 205, the signal that spectrum analysis is completed carries out the correction of frequency, amplitude, and amplitude compensation; When adopt just FIR algorithm time, only perform step S205, the signal that spectrum analysis is completed carries out amplitude compensation; When adopt just fft algorithm time, only perform step S204, the signal that spectrum analysis is completed carries out the correction of frequency, amplitude.

Certainly, in reality is implemented, when carrying out spectrum analysis, also can only adopt other algorithms outside FIR algorithm and fft algorithm, when the result of spectrum analysis also has the problem of band logical fluctuate large and transitional zone decay, perform step S205; When the result of spectrum analysis also has spectrum leakage phenomenon, perform step S204.

It should be noted that, in actual enforcement, the method portable in the application, to embedded software or DSP (digital signal processor, digital signal processor), is inputted original signal in this embedded software or DSP and is realized the method.

To sum up, the present embodiment provides a kind of signal processing method, comprising: according to default conditioning condition, original signal is nursed one's health, obtained simulating signal; Described simulating signal is carried out to A/D conversion, obtain digital signal; According to FIR algorithm and fft algorithm, described digital signal is carried out to spectrum analysis, obtain first signal; According to default correction algorithm, the frequency of described first signal and amplitude are carried out to correcting process, obtain secondary signal; According to default offset rule, the amplitude compensation to described secondary signal, obtains the 3rd signal and exports computing equipment to.Adopt the method, after digital signal being carried out to spectrum analysis according to FIR algorithm and fft algorithm and completing, proceed the correcting process of frequency and amplitude, and amplitude is compensated.Signal is carried out to correcting process, subdued the spectrum leakage phenomenon of fft algorithm, alleviated the fence effect of fft algorithm simultaneously; And signal is carried out to amplitude compensation, and made up the problem of the logical fluctuation of FIR band and transitional zone decay, improved the precision of signal processing results.

In specific implementation, this correction algorithm can adopt FFT difference correction algorithm.

As Fig. 3, the process flow diagram of a kind of signal processing method embodiment 2 that the application who illustrates provides, comprising:

Step S301: according to default conditioning condition, original signal is nursed one's health, obtained simulating signal;

Step S302: described simulating signal is carried out to A/D conversion, obtain digital signal;

Step S303: according to FIR algorithm and fft algorithm, described digital signal is carried out to spectrum analysis, obtain first signal;

Wherein, step S301-303 is consistent with the step S201-203 in embodiment 1, and the present embodiment does not repeat.

Step S304: the discrete spectrum that calculates described first signal according to default FFT interpolation algorithm;

Concrete, according to the first default sample frequency, first signal is sampled, obtain discrete signal; Utilize Fourier transform and default window function, the time-domain signal of this discrete signal is converted to frequency-region signal, obtain the continuous frequency spectrum of this first signal; This continuous frequency spectrum is continued to sampling and obtain some discrete spectral lines, this sample frequency is relevant to the first sample frequency.

For example, this first signal is frequency f ₀, amplitude is that A, phase angle are endless sine wave signal x (t), the expression formula of this first signal is:

With sample frequency f _safter x (t) sampling, obtain the discrete signal of following form:

The forms of time and space of the window function adopting is w (n), and its continuous frequency spectrum is W (2 π f),

$W\left(2\mathrm{\πf}\right)=\underset{n=-\∝}{\overset{n=+\∝}{\mathrm{\Σ}}}w\left(n\right)e^{-j2\mathrm{\πfn}}---\left(3\right)$

Utilizing Fourier transform, will be that the continuous frequency spectrum that obtains x (t) signal is with the red brill of signal with for frequency domain

Ignore the spectrum peak secondary lobe impact of negative frequency place, obtain

To this link frequency spectrum with sample frequency sample, wherein, N is data length, obtains

The formula that this sampling obtains (6) is some discrete spectral lines.

Step S305: the spectral line of selecting default number in described discrete spectrum;

Concrete, comprising: the spectral line of determining amplitude maximum in described discrete spectral line; Select remaining spectral line in the described default number adjacent with described maximum spectral line.

The process of wherein determining the spectral line of amplitude maximum in described discrete spectral line is specially: with described predeterminated position, be as the criterion, described discrete spectral line is adopted to the sequence of bubble sort method, obtain the spectral line of amplitude maximum.

Wherein, described default number comprises: 2,3 or 5.

Wherein, when being 2, the spectral line of choosing comprises: near time large spectral line spectral line of amplitude maximum and this spectral line; When being 3, the spectral line of choosing comprises: 2 adjacent spectral lines before and after the spectral line of amplitude maximum and this spectral line; When being 5, the spectral line of choosing comprises: 4 adjacent spectral lines before and after the spectral line of amplitude maximum and this spectral line, each two spectral lines of front and back.

In the present embodiment, take and choose 3 spectral lines and describe as example.

If k ₀possible position integer, near spectral line maximum it is k ₁, k ₁left side spectral line is k ₂, right side spectral line is k ₃, selection spectral line is k ₁, k ₂and k ₃.

Step S306: calculate the spectral line amplitude that each spectral line is corresponding;

Wherein, according to the spectral line of selecting in step S305, calculate the offset distance value of this spectral line and predeterminated position, and the discrete formula obtaining according to sampling calculates the amplitude corresponding to spectral line of selection.

For example, selecting spectral line is k ₁, k ₂and k ₃, the amplitude of each spectral line is respectively $y_2=\left|\stackrel{\‾}{X}\left(k_2\mathrm{\Δf}\right)\right|,y_3=\left|\stackrel{\‾}{X(}{k}_3\mathrm{\Δf}\right)|,$ Introduce parameter alpha=k ₀-k ₁.

Due to k ₁for k ₀near maximum spectral line, so the span of α is [0.5,0.5], has reacted k ₀depart from k ₁distance, obtain k ₁=k ₀-α, k ₂=k ₀-α-1, k ₂=k ₀-α+1, substitution formula (6)

$y_1=\stackrel{\‾}{X}\left(k_1\mathrm{\Δf}\right)=\frac{A}{2}W\left(2\mathrm{\π}\left(\frac{k_1\frac{f_s}{N}-f_0}{f_s}\right)\right)=\frac{A}{2}W\left(\frac{2\mathrm{\π}(k_1-k_0)}{N}\right)=\frac{A}{2}W\left(\frac{2\mathrm{\π}(-\mathrm{\α})}{N}\right)---\left(7\right)$

$y_2=\stackrel{\‾}{X}\left(k_2\mathrm{\Δf}\right)=\frac{A}{2}W\left(2\mathrm{\π}\left(\frac{k_2\frac{f_s}{N}-f_0}{f_s}\right)\right)=\frac{A}{2}W\left(\frac{2\mathrm{\π}(k_2-k_0)}{N}\right)=\frac{A}{2}W\left(\frac{2\mathrm{\π}(-\mathrm{\α}-1)}{N}\right)---\left(8\right)$

$y_3=\stackrel{\‾}{X}\left(k_3\mathrm{\Δf}\right)=\frac{A}{2}W\left(2\mathrm{\π}\left(\frac{k_3\frac{f_s}{N}-f_0}{f_s}\right)\right)=\frac{A}{2}W\left(\frac{2\mathrm{\π}(k_3-k_0)}{N}\right)=\frac{A}{2}W\left(\frac{2\mathrm{\π}(-\mathrm{\α}+1)}{N}\right)---\left(9\right)$

Step S307: the window function that foundation is default and described spectral line amplitude and default fitting condition calculate update information, and according to described update information, the frequency of described first signal and amplitude are revised, secondary signal obtained;

Wherein, according to the amplitude of above-mentioned spectral line and spectral line, in conjunction with default window function, calculate one group with pending with parameter group corresponding to signal, this parameter group and default fitting formula are carried out to matching, obtain the update information of amplitude, frequency, according to this update information, the amplitude of first signal and frequency are revised, obtained secondary signal.

As shown in Figure 4, the particular flow sheet of this step S307, comprising:

Step S401: construct respectively the first function and the second function according to described amplitude;

Example in continuation, according to spectral line k ₁, k ₂and k ₃amplitude constructed fuction 1 and function 2:

Function 1:

$\frac{y_2+y_1}{y_3+y_1}=\frac{\frac{A}{2}W\left(\frac{2\mathrm{\π}(-\mathrm{\α}-1)}{N}\right)+\frac{A}{2}W\left(\frac{2\mathrm{\π}(-\mathrm{\α})}{N}\right)}{\frac{A}{2}W\left(\frac{2\mathrm{\π}(-\mathrm{\α}+1)}{N}\right)+\frac{A}{2}W\left(\frac{2\mathrm{\π}(-\mathrm{\α})}{N}\right)}=g\left(\mathrm{\α}\right)---\left(10\right)$

Function 2:

$y_1+y_2+y_3=\frac{A}{2}(W\left(\frac{2\mathrm{\π}(-\mathrm{\α})}{N}\right)+W\left(\frac{2\mathrm{\π}(-\mathrm{\α}-1)}{N}\right)+W\left(\frac{2\mathrm{\π}(-\mathrm{\α}+1)}{N}\right))=\frac{A}{2}\mathrm{\ρ}\left(\mathrm{\α}\right)---\left(11\right)$

Step S402: according to the fitting function of described default window function, MATLAB, respectively to described the first function and the second Function Fitting, obtain the actual frequency value function of described first signal and the fitting polynomial formulas of actual magnitude function, and utilize described fitting polynomial formulas to carry out correcting process to the frequency of described first signal and amplitude.

Wherein, in the present embodiment, this window function adopts Blackman window function.

Order simulate the inverse function of g (x),

α＝g ^-1(β)＝a ₀+a ₁β+a ₂β ²+...+a _nβ ⁿ (12)

Wherein, a ₀, a ₁, a ₂... a _ncan value arbitrary real number.

Specifically can utilize the inverse function of the relevant polynomial interpolation instrument matching g (x) of MATLAB (MATrix LABoratory, matrix experiment chamber).

Hence one can see that, and the actual frequency values of first signal is

f ₀＝k ₀Δf＝(k ₁+α)Δf (13)

In like manner, utilize polynomial interpolation to simulate the inverse function of ρ (x) in formula (11),

A＝N ^-1(y ₁+y ₂+y ₃)ρ ^-1(α)＝N ^-1(y ₁+y ₂+y ₃)(b ₀+b ₁α+b ₂α ²+...+b _nα ⁿ) (14)

Wherein, b ₀, b ₁, b ₂... b _ncan value arbitrary real number.

Wherein, the time-domain expression of Blackman window function is

$w\left(n\right)=0.42-0.5\cos \left(\frac{2\mathrm{\πn}}{N}\right)+0.08\left(\frac{2\mathrm{\π}}{N}2n\right)---\left(15\right)$

Because the frequency of processed first signal is consistent, α can calculate, and β also can calculate, in like manner, also can calculate.

Change sample frequency, as f=f+0.001, repeat in above-mentioned steps x (t) is sampled, and after the computation process that needs, finally draw one group of α, β, γ value.

Utilize the fitting function of MATLAB to adopt polyfit, xx=polyfitg ^-1(β, α, 3), yy=polyfitg ^-1(alpha, gamma, 3).

Fitting of a polynomial function p=polyfitg ^-1in (x, y, n), the top step number of n representative polynomial, x, y be will matching data, p is the multinomial coefficient p=[p that matching obtains ₁, p ₂..., p _n, p _n+1], and meet y=p ₁x ⁿ+ p ₂x ^n-1+ ...+p _nx+p _n+1.

In conjunction with above-mentioned fitting function, obtain the fitting polynomial formulas (the matching amount of frequency values) of formula (12)

α＝1.9458-2.6332β+0.7089β ²-0.0221β ³ (16)

In conjunction with above-mentioned fitting function, obtain the fitting polynomial formulas (the matching amount of amplitude) of formula (14)

A＝N ^-1(y ₁+y ₂+y ₃)(2.1770+0.0006α+0.5631α ²-0.0105α ³) (17)

According to the matching amount in formula (16) and (17), to the amplitude A in first signal and frequency f ₀compensate, obtain secondary signal.

It should be noted that, the mode that has adopted polynomial expression to approach in the present embodiment, FFT correction algorithm data calculated amount is little, has simplified computation process.

And, in compensating process, adopted the method for sampling, reduced the expense of system.

Step S308: according to default offset rule, the amplitude compensation to described secondary signal, obtains the 3rd signal and export computing equipment to.

Wherein, step S308 is consistent with the step S205 in embodiment 1, and the present embodiment does not repeat.

It should be noted that, what the window function in the present embodiment adopted is Blackman window function, certainly, is not limited thereto, and in actual enforcement, also can adopt rectangular window function, Hanning window function or hamming window function, in the present embodiment, explains no longer in detail.

To sum up, a kind of information processing method that the present embodiment provides, adopts FFT difference correction algorithm to revise the amplitude of first signal and frequency, obtains secondary signal.Due to this FFT difference correction algorithm, with preceding step that described digital signal is carried out to the fft algorithm of spectrum analysis employing is corresponding, the problem of the spectrum leakage of appearance when adopting fft algorithm to carry out spectrum analysis that can be pointed.

In specific implementation, this can adopt FIR compensating coefficient table to compensate the amplitude of signal.

As Fig. 5, the process flow diagram of a kind of signal processing method embodiment 3 that the application who illustrates provides, comprising:

Step S501: according to default conditioning condition, original signal is nursed one's health, obtained simulating signal;

Step S502: described simulating signal is carried out to A/D conversion, obtain digital signal;

Step S503: according to FIR algorithm and fft algorithm, described digital signal is carried out to spectrum analysis, obtain first signal;

Step S504: according to default correction algorithm, the frequency of described first signal and amplitude are carried out to correcting process, obtain secondary signal;

Wherein, step S501-504 is consistent with the step S201-204 in embodiment 1, and the present embodiment does not repeat.

Step S505: search the compensated information corresponding with the frequency of described secondary signal from default FIR compensating coefficient table;

Wherein, the corresponding relation that comprises frequency and attenuation coefficient in this FIR compensating coefficient table.

Concrete, according to the frequency of this secondary signal, from this FIR compensating coefficient table, search corresponding compensated information, i.e. attenuation coefficient.

It should be noted that, this compensating coefficient table can be arranged in computing equipment for unified, also can be and is preset in other memory devices, in the present embodiment, does not limit.

Step S506: to the compensation of tabling look-up of the amplitude of described secondary signal, obtain the 3rd signal and export computing equipment to according to described compensated information.

Wherein, according to the compensated information finding and attenuation coefficient, the amplitude of secondary signal is compensated, to make up, adopt FIR algorithm to carry out the logical large and transitional zone decay of fluctuating of band that bandpass filtering brings.

Introduce the process of establishing of compensation meter below.

Take sample frequency as 500Hz, exponent number as 128,2～12Hz bandpass filtering is elaborated as example.

Utilize MATLAB software to generate one group of filter factor according to above-mentioned requirements; Utilize MATLAB to draw the frequency response chart that this classification coefficient is corresponding, this response diagram is as Fig. 6; By normalized frequency, convert Fig. 6 medium frequency response transverse axis data to actual frequency, and only pay close attention to transitional zone and the logical interior frequency component of band, longitudinal axis data-switching becomes attenuation coefficient, obtains the graph of a relation between the frequency shown in Fig. 7 and amplitude attenuation coefficient; Observation Fig. 7 is known, and 0Hz component decays to 0.5 times of original component, and 7Hz component decays to 1 times of original component, and 14Hz component decays to 0.25 times of original component.Obvious, this filter transmission band stopband fluctuation is large, if do not process, is difficult to meet requirement of engineering.Said frequencies response change is become to the form of form, so that complete FFT, calculate and to finish rear contrast frequency values and carry out the correction of amplitude.

To sum up, a kind of information processing method that the present embodiment provides, adopts FIR compensating coefficient table to compensate the amplitude of signal, obtains the 3rd signal.This FIR compensating coefficient table is default form, when compensating, only need be according to the frequency of signal to be compensated, from this compensation meter, search, data calculated amount is little, and this FIR compensating coefficient table, with preceding step that described digital signal is carried out to the FIR algorithm of spectrum analysis employing is corresponding, the problem that the logical fluctuation of band is large and transitional zone decays of appearance when adopting FIR algorithm to carry out spectrum analysis that can be pointed.

A kind of signal processing method embodiment providing with above-mentioned the application is corresponding, and the application also provides a kind of signal processing apparatus embodiment.

As Fig. 8, show the structural representation of a kind of signal processing apparatus embodiment 1 that the application provides, comprising: conditioning module 801, modular converter 802, spectrum analysis module 803, correcting module 804 and compensating module 805;

Conditioning module 801, for according to default conditioning condition, original signal being nursed one's health, obtains simulating signal;

Wherein, the conditioning condition of this conditioning module 801 can be the condition of follow-up A/D transformational structure input signal.

Concrete, this conditioning process comprises: power amplification, filtering, linear compensation, isolation and/or protection etc., and the process of conditioning can be set according to actual conditions, and the present embodiment does not repeat.

Modular converter 802, for described simulating signal is carried out to A/D conversion, obtains digital signal;

Wherein, because follow-up FIR algorithm and fft algorithm need to be according to digital format signals, and what after conditioning, obtain is analog format signal, and therefore, 802 pairs of these simulating signals of modular converter are carried out A/D conversion, make this signal be converted to simulating signal.

Spectrum analysis module 803, for according to FIR algorithm and fft algorithm, described digital signal being carried out to spectrum analysis, obtains first signal;

Wherein, spectrum analysis module 803 is carried out spectrum analysis to digital signal successively according to FIR algorithm and fft algorithm, obtains first signal.

Concrete, these spectrum analysis module 803 spectrum analysis processes are: adopt fft algorithm to realize the conversion to digital signal, adopt FIR algorithm to realize the filtering to digital signal.

But, adopt fft algorithm, in this conversion process, easily there is spectrum leakage phenomenon; And there is the logical problem large and transitional zone decay that fluctuates of band in the filtering that adopts FIR algorithm.

Therefore, at follow-up correcting module 804 and compensating module 805, so that the problems referred to above are made up.

Correcting module 804, for according to default correction algorithm, the frequency of described first signal and amplitude being carried out to correcting process, obtains secondary signal;

Wherein, before carrying out signal processing, to the default correction algorithm of correcting module 804.

Concrete, correcting module 804, according to this correction algorithm, is revised respectively the frequency of first signal and amplitude, the problem of the spectrum leakage of appearance while carrying out digital conversion to overcome fft algorithm.

Wherein, the 3rd signal is exported to follow-up computing equipment, to carry out follow-up computation process.

Wherein, described correction algorithm comprises FFT difference correction algorithm, this correcting module 804 specifically for: according to default FFT interpolation algorithm, calculate the discrete spectrum of described secondary signal; Select the spectral line of default number in described discrete spectrum; Calculate the offset distance value of each spectral line and predeterminated position, and spectral line amplitude corresponding to described offset distance value; The window function that foundation is default and described spectral line amplitude and default fitting condition calculate update information, and according to described update information, the frequency of described first signal and amplitude are revised, and obtain secondary signal.

Concrete makeover process describes in above-mentioned embodiment of the method, and the present embodiment does not repeat.

Compensating module 805, for the default offset rule of foundation, the amplitude compensation to described secondary signal, obtains the 3rd signal and exports computing equipment to.

Wherein, before carrying out signal processing, to the default offset rule of compensating module 805.

Concrete, compensating module 805 is according to this offset rule, to carrying out amplitude compensation through the secondary signal of correction, to overcome the problem that the logical fluctuation of the band that adopts the filtering of FIR algorithm to cause is large and transitional zone decays.

It should be noted that, while digital signal being carried out to spectrum analysis in the application, what adopt is FIR algorithm and fft algorithm, in actual enforcement, when carrying out spectrum analysis, also can only adopt a kind of in FIR algorithm and fft algorithm, accordingly, when adopt be above-mentioned two kinds of algorithms time, triggering correcting module 804 and compensating module 805, the signal that spectrum analysis is completed carries out the correction of frequency, amplitude, and amplitude compensation; When adopt just FIR algorithm time, only trigger compensating module 805, the signal that spectrum analysis is completed carries out amplitude compensation; When adopt just fft algorithm time, only trigger correcting module 804, the signal that spectrum analysis is completed carries out the correction of frequency, amplitude.

Certainly, in reality is implemented, when carrying out spectrum analysis, also can only adopt other algorithms outside FIR algorithm and fft algorithm, when the result of spectrum analysis also has the problem of band logical fluctuate large and transitional zone decay, trigger compensating module 805; When the result of spectrum analysis also has spectrum leakage phenomenon, trigger correcting module 804.

It should be noted that, in actual enforcement, the method portable in the application, to embedded software or DSP, is inputted original signal in this embedded software or DSP and is realized the method.

To sum up, the present embodiment provides a kind of signal processing apparatus, comprising: conditioning module, for according to default conditioning condition, original signal being nursed one's health, obtains simulating signal; Modular converter, for described simulating signal is carried out to A/D conversion, obtains digital signal; Spectrum analysis module, for according to FIR algorithm and fft algorithm, described digital signal being carried out to spectrum analysis, obtains first signal; Correcting module, for according to default correction algorithm, the frequency of described first signal and amplitude being carried out to correcting process, obtains secondary signal; Compensating module, for the default offset rule of foundation, the amplitude compensation to described secondary signal, obtains the 3rd signal and exports computing equipment to.。Adopt this device, after digital signal being carried out to spectrum analysis according to FIR algorithm and fft algorithm and completing, proceed the correcting process of frequency and amplitude, and amplitude is compensated.Signal is carried out to correcting process, subdued the spectrum leakage phenomenon of fft algorithm, alleviated the fence effect of fft algorithm simultaneously; And signal is carried out to amplitude compensation, and made up the problem of the logical fluctuation of FIR band and transitional zone decay, improved the precision of signal processing results.

Claims (10)

1. a signal processing method, is characterized in that, comprising:

According to default conditioning condition, original signal is nursed one's health, obtained simulating signal;

Described simulating signal is carried out to analog/digital A/D conversion, obtain digital signal;

According to having limit for length's impulse response FIR algorithm and fast fourier transform fft algorithm to carry out spectrum analysis to described digital signal, obtain first signal;

According to default correction algorithm, the frequency of described first signal and amplitude are carried out to correcting process, obtain secondary signal;

According to default offset rule, the amplitude compensation to described secondary signal, obtains the 3rd signal and exports computing equipment to.

2. method according to claim 1, is characterized in that, described correction algorithm comprises FFT difference correction algorithm, and the default correction algorithm of described foundation carries out correcting process to the frequency of described first signal and amplitude and specifically comprises:

According to default FFT interpolation algorithm, calculate the discrete spectrum of described secondary signal;

Select the spectral line of default number in described discrete spectrum;

Calculate the spectral line amplitude that each spectral line is corresponding;

The window function that foundation is default and described spectral line amplitude and default fitting condition calculate update information, and according to described update information, the frequency of described first signal and amplitude are revised, and obtain secondary signal.

3. method according to claim 1, is characterized in that, described foundation is preset offset rule, and the amplitude compensation of described secondary signal is specifically comprised:

From default FIR compensating coefficient table, search the compensated information corresponding with the frequency of described secondary signal;

According to described compensated information, to the compensation of tabling look-up of the amplitude of described secondary signal, obtain the 3rd signal.

4. method according to claim 2, is characterized in that, in the described discrete spectrum of described selection, the spectral line of default number comprises:

Determine the spectral line of amplitude maximum in described discrete spectral line;

Select remaining spectral line in the described default number adjacent with described maximum spectral line.

5. method according to claim 2, is characterized in that, the window function that described foundation is default and described spectral line amplitude, and default fitting condition, comprise the frequency of described first signal and amplitude correction:

According to described amplitude, construct respectively the first function and the second function;

Fitting function according to described default window function, matrix experiment chamber MATLAB, respectively to described the first function and the second Function Fitting, obtain the actual frequency value function of described first signal and the fitting polynomial formulas of actual magnitude function, and utilize described fitting polynomial formulas to carry out correcting process to the frequency of described first signal and amplitude.

6. method according to claim 2, is characterized in that, described window function adopts Blackman window function.

7. method according to claim 4, is characterized in that, in described definite described discrete spectral line, the spectral line of amplitude maximum comprises:

With described predeterminated position, be as the criterion, described discrete spectral line is adopted to the sequence of bubble sort method, obtain the spectral line of amplitude maximum.

8. method according to claim 2, is characterized in that, described default number comprises: 2,3 or 5.

9. a signal processing apparatus, is characterized in that, comprising:

Conditioning module, for according to default conditioning condition, original signal being nursed one's health, obtains simulating signal;

Modular converter, for described simulating signal is carried out to A/D conversion, obtains digital signal;

Spectrum analysis module, for according to FIR algorithm and fft algorithm, described digital signal being carried out to spectrum analysis, obtains first signal;

Correcting module, for according to default correction algorithm, the frequency of described first signal and amplitude being carried out to correcting process, obtains secondary signal;

Compensating module, for the default offset rule of foundation, the amplitude compensation to described secondary signal, obtains the 3rd signal and exports computing equipment to.

10. device according to claim 9, is characterized in that, described correction algorithm comprises FFT difference correction algorithm, described correcting module specifically for:

According to default FFT interpolation algorithm, calculate the discrete spectrum of described secondary signal;

Select the spectral line of default number in described discrete spectrum;

Calculate the spectral line amplitude that each spectral line is corresponding;

The window function that foundation is default and described spectral line amplitude and default fitting condition calculate update information, and according to described update information, the frequency of described first signal and amplitude are revised, and obtain secondary signal.