CN103412004A - Method for detecting storage time of citrus sinensis - Google Patents

Abstract

The invention discloses a method for detecting the storage time of citrus sinensis. According to the method, response signals of citrus sinensis samples with different storage time are measured by using a gas sensor and combining an acoustic surface wave; a first prediction model and a second prediction model of citrus sinensis storage time are established according to gas sensor array response data and acoustic surface wave detection frequency; a citrus sinensis storage time comprehensive prediction model is established. The method has the advantages of quickness, no loss and high accuracy; storage and picking time of the citrus sinensis can be arranged reasonably according to storage time; deterioration of the citrus sinensis in transportation and storage processes can be effectively avoided.

Description

The detection method in a kind of sweet orange storage time

Technical field

The present invention relates to fruit freshness detection field, especially relate to a kind ofly for the storage time quick, the Non-Destructive Testing sweet orange, and detect the detection method in the sweet orange storage time that degree of accuracy is high.

Background technology

Sweet orange (Citrus Sinensis) is the fruit of Rutaceae citrus plant orange tree, also is called yellow fruit, and the fruit circle is to Long Circle, orange-yellow, the oil vacuole projection, and pericarp is difficult for peeling off.Without bitter taste.Newel enriches, and the juice flavor is sweet and fragrant, and contains a large amount of sugar and a certain amount of citric acid and abundant vitamin C, and nutritive value is higher.

Because the skin of sweet orange is thinner, in storage and transport process, be subject to the impact that environmental factor changes and produce quality and descend even rotten.

Though the artificial sense evaluation method can be distinguished the slight change of sweet orange in storage process, the result of the method causes repeatability and referential all poor because estimating the factor impacts such as personnel's individual difference, health status.

Though the physical and chemical inspection method can be reacted the freshness situation of fruit sample, but experimental implementation is loaded down with trivial details, required time of a specified duration, be difficult to meet the needs that quick nondestructive detects.

The instrumental methods such as vapor-phase chromatography (GC) and gas chromatography-mass spectrography technology (GC-MS), though but the volatilization gas of the different storage life sweet orange samples of Accurate Analysis becomes grading information, but this detection method testing cost is expensive, sense cycle is long, and the gained odour component is all the product of sample after separating.In addition, instrumental analysis has larger dependence to operating personnel's skilled operation degree usually.

Therefore, in the urgent need to the novel detection technique in a kind of judgement orange storage time that can quick nondestructive.

Chinese patent mandate publication number: CN103163217A, authorize open day on June 19th, 2013, a kind of SAW (Surface Acoustic Wave) resonator Series detectors system and detection method thereof are disclosed, described detection system comprises resonance detecting device, direct supply, digital counter, digital counter is connected with computing machine by the RS232 communication interface, the resonance detecting device is arranged in shielded box, and this resonance detecting device has two probes, between two probes be used to placing manaca to be detected; The method of using described detection system to detect the manaca storage life comprises the following steps: step 1: place manaca to be measured between two probes of detection system; Step 2: manaca to be measured is measured, gathered the load frequency of resonance detecting device by digital counter; Step 3: read by computing machine the load frequency data that digital counter gathers, by calculating the manaca storage time.Weak point is that this invention has the deficiency of the poor accuracy of detection.

Summary of the invention

The present invention is for the deficiency of the poor accuracy that overcomes detection method detection of the prior art, provides a kind of for the storage time quick, the Non-Destructive Testing sweet orange, and detects the detection method in the sweet orange storage time that degree of accuracy is high.

To achieve these goals, the present invention is by the following technical solutions:

The detection method in a kind of sweet orange storage time, comprise the steps:

(1-1) select n the sweet orange sample of just having plucked, the sweet orange sample be placed in congealer and store, the sweet orange sample is detected:

(1-1-1) setting detection time is m ₁, the sequence number in storage time is Time ₁, Time ₁Initial value be 1; Use acoustic surface wave detection device to detect the frequency response of sweet orange, and obtain sweet orange storage time first predictor formula:

Described gas-detecting device comprises air chamber, is located at sensor array and analog to digital converter in air chamber; Described air chamber is provided with sampling probe and cleans probe, and sampling probe is provided with the sampling air pump, cleans probe and is provided with the cleaning air pump; Sensor array is electrically connected to analog to digital converter, is equipped with the data-interface be electrically connected to for computing machine on analog to digital converter, sampling air pump and cleaning air pump; Described sensor array comprises several gas sensors;

Step a, optional 1 sweet orange sample in n sweet orange sample, and the sweet orange sample is put into to shielded box, two electrodes of SAW (Surface Acoustic Wave) resonator are contacted with the opposite flank of sweet orange sample respectively;

Step b, acoustic surface wave detection device work is after 40 to 60 minutes, and counter gathers the frequency response curve of oscillation circuit, on frequency response curve, gathers 80 to 110 frequency values, and each frequency values is formed to frequency signal Input (t);

In computing machine, be preset with the first stochastic resonance system model Wherein, ξ (t) is white Gaussian noise, and A, a and b are all constants, f ₀Be frequency modulating signal, D is noise intensity, and t is the Brownian movement Particles Moving time, and x is the coordinate of Particles Moving;

By in frequency signal Input (t) input the first stochastic resonance system model, the first stochastic resonance system model is resonated;

Computing machine utilizes formula $\mathrm{SNR}=2\left[\underset{\mathrm{\Δ\ω}\→0}{\lim }\underset{\mathrm{\Ω}-\mathrm{\Δ\ω}}{\overset{\mathrm{\Ω}+\mathrm{\Δ\ω}}{\∫}}S\left(\mathrm{\ω}\right)d\mathrm{\ω}\right]/S_N\left(\mathrm{\Ω}\right)$ Calculate output signal-to-noise ratio SNR; Wherein, ω is signal frequency, and Ω is angular frequency, and S (ω) is signal spectral density, S _N(Ω) be the noise intensity in signal frequency range;

Computing machine draws the output signal-to-noise ratio curve in the first stochastic resonance system model, obtains two peak values of signal to noise ratio (S/N ratio) curve, using the absolute value of the difference of two peak values as signal to noise ratio (S/N ratio) eigenwert F; And signal to noise ratio (S/N ratio) eigenwert F is stored in computing machine;

Step c, work as Time ₁<m ₁, make Time ₁Value increases by 1, and will be detected the sweet orange sample and put back in congealer and store, after 24 hours, repeating step a and step b;

Obtain m ₁Individual and Time ₁The F be associated, by F and Time ₁Form point (F, Time ₁), according to m ₁Individual point (F, Time ₁) carry out linear fit, obtain sweet orange storage time first predictor formula: Time ₁=f (F);

Sweet orange storage time first predictor formula is relevant with the sweet orange storage time, the first predictor formula difference in different storage times;

(1-1-2) setting detection time is m ₂, the sequence number in storage time is Time ₂, Time ₂Initial value be 1; Use gas-detecting device to detect the smell of sweet orange volatilization, and obtain sweet orange storage time second predictor formula:

Described gas-detecting device comprises air chamber, is located at sensor array and analog to digital converter in air chamber; Described air chamber is provided with sampling probe and cleans probe, and sampling probe is provided with the sampling air pump, cleans probe and is provided with the cleaning air pump; Sensor array is electrically connected to analog to digital converter, is equipped with the data-interface be electrically connected to for computing machine on analog to digital converter, sampling air pump and cleaning air pump; Described sensor array comprises several gas sensors;

Steps d, computer control cleaning gas pump work, pure air passes in air chamber each gas sensor is cleaned, and when the response of each gas sensor is stablized to baseline, closes the cleaning air pump;

Step e, optional 1 sweet orange sample, inserted the sweet orange sample in sample bottle in n sweet orange sample, by sample bottle sealing and standing 30 to 40 minutes, sampling probe and air pressure balancer are injected in sample bottle simultaneously to the escaping gas that computer control sampling probe collected specimens produces; When gathering escaping gas, air pressure balancer will import in sample bottle through the air of activated carbon filtration, realizes air pressure balance;

Step f, escaping gas contacts with each sensor, and each sensor produces respectively analog response signal; Analog to digital converter is converted to respectively the digital response signal by each analog response signal, and computing machine is averaged digital response signal, obtains sensor array digital response signal I (t):

In computing machine, be preset with the second accidental resonance model

Wherein, ξ (t) is white Gaussian noise, and t is the Brownian movement Particles Moving time, and x is the coordinate of Particles Moving, f ₀Be signal frequency, a, b and A are constant;

Step g, by I (t) input the second accidental resonance model, make the second stochastic resonance system model produce accidental resonance;

Computing machine utilizes formula $\mathrm{SNR}=2\left[\underset{\mathrm{\&Delta;\&omega;}\&RightArrow;0}{\lim }\underset{\mathrm{\&Omega;}-\mathrm{\&Delta;\&omega;}}{\overset{\mathrm{\&Omega;}+\mathrm{\&Delta;\&omega;}}{\&Integral;}}S\left(\mathrm{\&omega;}\right)d\mathrm{\&omega;}\right]/S_N\left(\mathrm{\&Omega;}\right)$ Calculate output signal-to-noise ratio SNR; Wherein, ω is signal frequency, and Ω is angular frequency, and S (ω) is signal spectral density, S _N(Ω) be the noise intensity in signal frequency range;

Step h, computing machine draw the output signal-to-noise ratio curve of the second stochastic resonance system model, and the signal to noise ratio (S/N ratio) maximal value of choosing the signal to noise ratio (S/N ratio) curve in the signal to noise ratio (S/N ratio) curve is eigenwert SNR _max, and by SNR _maxBe stored in computing machine;

Step I, work as Time ₂<m ₂, make Time ₂Value increases by 1, will be detected the sweet orange sample and put back in congealer and store, and after 24 hours, repeating step d is to step h;

Obtain m ₂Individual and Time ₂The SNR be associated _max, by SNR _maxAnd Time ₂Form point (SNR _max, Time ₂), according to m point (SNR _max, Time ₂) carry out the Nonlinear Boltzmann match, obtain sweet orange storage time second predictor formula Time ₂=f (SNR _max);

Sweet orange storage time second predictor formula is relevant with the sweet orange storage time, the second predictor formula difference in different storage times;

(1-2) set up sweet orange storage time Comprehensive Model: Time (F, SNR _max)=P ₁* f (F)+P ₂* f (SNR _max), P ₁And P ₂For undetermined coefficient; Two groups of sweet orange storage times, the signal to noise ratio (S/N ratio) eigenwert F corresponding with the described storage time are reached to the SNR corresponding with the described storage time _maxSubstitution sweet orange storage time Comprehensive Model, thus P obtained ₁And P ₂

(1-3) detect with described sweet orange sample same stored condition under sweet orange sample W to be detected, obtain the signal to noise ratio (S/N ratio) maximal value SNR of sweet orange sample W _maxWith signal to noise ratio (S/N ratio) eigenwert F; By SNR _maxIn frequency F substitution sweet orange storage time Comprehensive Model, calculate the storage time Time of sweet orange sample W _Prediction.

Surface acoustic wave (SAW) technology is born in the seventies in last century, and the method is that manufactured size is little, the checkout equipment of high sensitivity, high integration provides the foundation.At present, existing many application reports of detection based on SAW (Surface Acoustic Wave) device.At food and field of biological detection, also be widely used, as the detection of bacterium in milk, bacterial growth condition monitoring, the detection of pancreas lipase, bio-medical analysis etc.For surface acoustic wave sensing detection technology, the focus studied for a long period of time concentrates on the chemical/biological on SAW (Surface Acoustic Wave) resonator surface and modifies, in testing process if there is specific reaction (as the antigen antagonist, acceptor is to specific effect processes such as parts), when now surface acoustic wave is by the resonator piezoelectric substrate, respective change occurs in ripple information (velocity of wave etc.), thereby realizes that measured object kind/concentration information characterizes.

The gas sensor detection technique is a kind of smell fingerprint detection method, the process of simulation people's olfactory system.The gas of take is analytic target, by the olfactory system of simulating the people, in real time the aromatic substance of ad-hoc location is caught and detects, thereby obtain characteristic signal, and therefore this smell finger print detection device is called Electronic Nose visually.Due to the function of its uniqueness, in food, cosmetics, petrochemical complex, wrappage, environmental monitoring, the field such as clinical, chemical, be widely applied, received the increasing concern of each side.

The accidental resonance technology is shown up prominently in detection data feature values extraction field at present.This theory is proposed in 1981 by Italian physicist Benzi, in order to the phenomenon of explaining that the earth meteorological glacial epoch in time immemorial and cycle warm climate phase alternately occur.Accidental resonance has three elements: nonlinear system, weak signal and noise source.From the signal process angle, consider, accidental resonance is in the nonlinear properties transmitting procedure, by regulating intensity or other parameter of system of noise, make system output reach optimum value, in fact also can think the collaborative state of input signal, nonlinear system, noise.

Generally, under the excitation of excitation noise, system produces accidental resonance, and now output signal is greater than input signal, thereby has played the effect that signal amplifies.Simultaneously, accidental resonance is transformed into the noise energy in the part detection signal in signal and goes, thereby has effectively suppressed the noisiness in the detection signal.Therefore, stochastic resonance system is equivalent to improve the effect of output signal-noise ratio, and signal, excitation noise and bistable system can be regarded an efficient signal processor as.On above technical foundation, stochastic resonance system output signal-to-noise ratio analytical technology can be reacted the essential characteristic information of sample preferably.

In in the frequency signal Input (t) that surface acoustic wave detects, comprising, grasp noise, by the accidental resonance analysis, in erasure signal, grasp noise, make testing result more stable, error is less, and accuracy is higher.

After accidental resonance was analyzed, between the sweet orange sample in different storage times, discrimination was better, and detection model linear fit precision is higher.

What accidental resonance output signal-to-noise ratio characteristic information reflected is the essential information of sample, and this characteristic information does not change with the restriction of detection method or multiplicity, only relevant with the character of sample, is conducive to the demarcation of properties of samples, improves accuracy of detection.

Accidental resonance analytical approach favorable reproducibility, repeat 100 times and calculate, and the resultant error ratio of output is no more than 0.1%.And the error ratio of the frequency signal error rate that simple surface acoustic wave detects after than accidental resonance Analysis signal-to-noise ratio (SNR) exceeds several times.

The present invention adopts gas sensor and in conjunction with surface acoustic wave, measures the response signal of different storage time sweet orange samples, according to gas sensor array response data and surface acoustic wave, detect frequency structure first, second forecast model of sweet orange storage time, and set up sweet orange storage time Comprehensive Model, and utilize sweet orange storage time Comprehensive Model to detect the storage time of sweet orange sample W.

What surface acoustic wave detected reflection is the fruit internal structural information, is the smell information of fruit outside and gas detects reflection, and this combination can accurately characterize the situation of change in the sweet orange storage process.The method is conducive to instruct such fruit timely collecting, the decline of as far as possible avoiding fruits nutrition that harvest in advance causes to be worth, and also can reducing gathers evening causes the corrupt loss brought.

And method of the present invention has advantages of fast, can't harm, accuracy is good, can to storage, the plucking time of sweet orange, carry out reasonable arrangement according to the storage time, effectively prevent the putrid and deteriorated of in transportation and storage process sweet orange.

As preferably, the centre frequency of described SAW (Surface Acoustic Wave) resonator is 433.92MHz.

As preferably, described n is 6 to 15.

As preferably, described m ₁And m ₂Be 10 to 180.

As preferably, described gas sensor is 8, be respectively the first sensor for detection of sulfide gas, for detection of the second sensor of inflammable gas, for detection of the 3rd sensor of Ammonia gas, for detection of the four-sensor of ethanol class gas, the 5th sensor for detection of hydrocarbon component gas, for detection of the 6th sensor of alkanes gas, for detection of the 7th sensor of propane and butane, for detection of the 8th sensor of nitride gas.

As preferably, described SAW (Surface Acoustic Wave) resonator comprises piezoelectric substrate, interdigital transducer, 2 reflecting gratings and is located at 4 gain grid of reflecting grating both sides; Described electrode is electrically connected to the link on being located at interdigital transducer.

Arranging of grid of gain can prevent that surface acoustic wave from overflowing from the blank position of reflecting grating, can make surface acoustic wave be reflected back interdigital transducer through reflecting grating, strengthens the intensity of surface acoustic wave.The width of gain grid is identical with reflecting grating, with reflecting grating apart from being 9 μ m.

As preferably, in described step c, pure air passes in air chamber each gas sensor was cleaned 40 to 50 minutes.

As preferably, in described step c, pure air cleans sensor with the flow velocity of 860mL/min to 1100mL/min.

Therefore, the present invention has following beneficial effect: (1) fast, harmless, accuracy is good; (2) effectively prevent sweet orange putrid and deteriorated in transportation and storage process.

The accompanying drawing explanation

Fig. 1 is a kind of process flow diagram of embodiments of the invention;

Fig. 2 is a kind of theory diagram of acoustic surface wave detection device of the present invention;

Fig. 3 is a kind of theory diagram of gas-detecting device of the present invention;

Fig. 4 is a kind of structural representation of SAW (Surface Acoustic Wave) resonator of the present invention;

Fig. 5 is the output signal-to-noise ratio curve in the first stochastic resonance system model of the present invention;

Fig. 6 is the matched curve of signal to noise ratio (S/N ratio) eigenwert F of the present invention;

Fig. 7 is signal to noise ratio (S/N ratio) eigenwert SNR of the present invention _maxMatched curve.

In figure: counter 1, shielded box 2, oscillator 3, SAW (Surface Acoustic Wave) resonator 4, piezoelectric substrate 5, interdigital transducer 6, reflecting grating 7, gain grid 8, computing machine 9, air chamber 10, sensor array 11, analog to digital converter 12, sampling probe 13, cleaning probe 14, sampling air pump 15, cleaning air pump 16, sweet orange 17, link 18, wireless launcher 19.

Embodiment

The present invention will be further described below in conjunction with the drawings and specific embodiments.

Embodiment as shown in Figure 1 is the detection method in a kind of sweet orange storage time, comprises the steps:

Step 100, select the sweet orange sample of the firm harvesting that 15 sizes are identical, and the sweet orange sample is placed in congealer and stores, and the refrigerated storage temperature of congealer is 4 ℃, and the sweet orange sample is carried out to following detection:

Step 101, setting detection time is m ₁=11, the sequence number in storage time is Time ₁, Time ₁Initial value be 1; Use acoustic surface wave detection device to detect the frequency response of sweet orange, and obtain sweet orange storage time first predictor formula:

As shown in Figure 2, acoustic surface wave detection device comprises counter 1, and shielded box 2 is located at oscillator 3 and SAW (Surface Acoustic Wave) resonator 4 in shielded box; Oscillator and SAW (Surface Acoustic Wave) resonator form oscillation circuit, and counter is electrically connected to oscillation circuit, and counter is provided with for the wireless launcher 19 with computing machine 9 wireless connections, and SAW (Surface Acoustic Wave) resonator is provided with two electrodes;

As shown in Figure 4, SAW (Surface Acoustic Wave) resonator comprises piezoelectric substrate 5,6,2 reflecting gratings 7 of interdigital transducer and is located at 4 of reflecting grating both sides gain grid 8; Electrode is electrically connected to the link 18 on being located at interdigital transducer.

Step a, optional 1 sweet orange sample in 15 sweet orange samples, and the sweet orange sample is put into to shielded box, two electrodes of SAW (Surface Acoustic Wave) resonator are contacted with the opposite flank of sweet orange 17 samples respectively;

Step b, acoustic surface wave detection device work is after 40 minutes, and counter gathers the frequency response curve of oscillation circuit, on frequency response curve, gathers 100 frequency values that are spacedly distributed, and each frequency values is formed to frequency signal Input (t);

In computing machine, be preset with the first stochastic resonance system model Wherein, ξ (t) is white Gaussian noise, and A, a and b are all constants, f ₀Be frequency modulating signal, D is noise intensity, and t is the Brownian movement Particles Moving time, and x is the coordinate of Particles Moving;

By in frequency signal Input (t) input the first stochastic resonance system model, the first stochastic resonance system model is resonated;

Computing machine utilizes formula $\mathrm{SNR}=2\left[\underset{\mathrm{\&Delta;\&omega;}\&RightArrow;0}{\lim }\underset{\mathrm{\&Omega;}-\mathrm{\&Delta;\&omega;}}{\overset{\mathrm{\&Omega;}+\mathrm{\&Delta;\&omega;}}{\&Integral;}}S\left(\mathrm{\&omega;}\right)\mathrm{d\&omega;}\right]/S_N\left(\mathrm{\&Omega;}\right)$ Calculate output signal-to-noise ratio SNR; Wherein, ω is signal frequency, and Ω is angular frequency, and S (ω) is signal spectral density, S _N(Ω) be the noise intensity in signal frequency range;

Computing machine draws the output signal-to-noise ratio curve in the first stochastic resonance system model, obtains two peak values of signal to noise ratio (S/N ratio) curve, using the absolute value of the difference of two peak values as signal to noise ratio (S/N ratio) eigenwert F; And signal to noise ratio (S/N ratio) eigenwert F is stored in computing machine;

For example, in output signal-to-noise ratio curve as shown in Figure 5 two peak values be respectively-70dB and-94dB, signal to noise ratio (S/N ratio) eigenwert F=|-94+70|=24dB.

Step c, work as Time ₁<11, make Time ₁Value increases by 1, and will be detected the sweet orange sample and put back in congealer and store, after 24 hours, repeating step a and step b;

Obtain 11 and Time ₁The F be associated, by F and Time ₁Form point (F, Time ₁), according to m ₁Individual point (F, Time ₁) carry out nonlinear fitting, obtain signal to noise ratio (S/N ratio) eigenwert matched curve shown in Figure 6 $F=4.00024+3.70916\&times;e^{{\mathrm{Time}}_1/3.89102},$ And obtain sweet orange storage time first predictor formula: ${\mathrm{Time}}_1=3.89102\ln \frac{F-4.00024}{3.70916};$

Step 102, setting detection time is m ₂=11, the sequence number in storage time is Time ₂, Time ₂Initial value be 1; Use gas-detecting device to detect the smell of sweet orange volatilization, and obtain sweet orange storage time second predictor formula:

As shown in Figure 3, gas-detecting device comprises air chamber 10, is located at sensor array 11 and analog to digital converter 12 in air chamber; Air chamber is provided with sampling probe 13 and cleans probe 14, and sampling probe is provided with sampling air pump 15, and the cleaning probe is provided with and cleans air pump 16; Sensor array is electrically connected to analog to digital converter, is equipped with the data-interface be electrically connected to for computing machine on analog to digital converter, sampling air pump and cleaning air pump; Sensor array comprises 8 gas sensors.

8 gas sensors are respectively the first sensor for detection of sulfide gas, the second sensor for detection of inflammable gas, the 3rd sensor for detection of Ammonia gas, four-sensor for detection of ethanol class gas, the 5th sensor for detection of hydrocarbon component gas, for detection of the 6th sensor of alkanes gas, for detection of the 7th sensor of propane and butane, for detection of the 8th sensor of nitride gas.

Steps d, computer control cleaning gas pump work, pure air pass in air chamber each gas sensor were cleaned 40 minutes, when the response of each gas sensor is stablized to baseline, close the cleaning air pump;

Step e, optional 1 sweet orange sample, inserted the sweet orange sample in sample bottle in 15 sweet orange samples, by sample bottle sealing and standing 30 minutes, sampling probe and air pressure balancer are injected in sample bottle simultaneously to the escaping gas that computer control sampling probe collected specimens produces; When gathering escaping gas, air pressure balancer will import in sample bottle through the air of activated carbon filtration, realizes air pressure balance;

Step f, escaping gas contacts with each sensor, and each sensor produces respectively analog response signal; Analog to digital converter is converted to respectively the digital response signal by each analog response signal, and computing machine is averaged digital response signal, obtains sensor array digital response signal I (t);

In computing machine, be preset with the second accidental resonance model

Wherein, ξ (t) is white Gaussian noise, and t is the Brownian movement Particles Moving time, and x is the coordinate of Particles Moving, f ₀Be signal frequency, a, b and A are constant;

Step g, by I (t) input the second accidental resonance model, make the second stochastic resonance system model produce accidental resonance;

Computing machine utilizes formula $\mathrm{SNR}=2\left[\underset{\mathrm{\&Delta;\&omega;}\&RightArrow;0}{\lim }\underset{\mathrm{\&Omega;}-\mathrm{\&Delta;\&omega;}}{\overset{\mathrm{\&Omega;}+\mathrm{\&Delta;\&omega;}}{\&Integral;}}S\left(\mathrm{\&omega;}\right)\mathrm{d\&omega;}\right]/S_N\left(\mathrm{\&Omega;}\right)$ Calculate output signal-to-noise ratio SNR; Wherein, ω is signal frequency, and Ω is angular frequency, and s (ω) is signal spectral density, s _N(Ω) be the noise intensity in signal frequency range;

Step h, computing machine draw the output signal-to-noise ratio curve of the second stochastic resonance system model, and the signal to noise ratio (S/N ratio) maximal value of choosing the signal to noise ratio (S/N ratio) curve in the signal to noise ratio (S/N ratio) curve is eigenwert SNR _max, and by SNR _maxBe stored in computing machine;

Step I, work as Time ₂<11, make Time ₂Value increases by 1, will be detected the sweet orange sample and put back in congealer and store, and after 24 hours, repeating step d is to step h;

Obtain 11 and Time ₂The SNR be associated _max, by SNR _maxAnd Time ₂Form point (SNR _max, Time ₂), according to m point (SNR _max, Time ₂) carry out the Nonlinear Boltzmann match, obtain matched curve as shown in Figure 7 ${\mathrm{SNR}}_{\max }=\frac{1.43759}{\left({1+e}^{\frac{{\mathrm{Time}}_2-11.53884}{1.58287}}\right)}-72.91711,$ And obtain sweet orange storage time second predictor formula ${\mathrm{Time}}_{}$ ${}_2=1.58287\ln (\frac{1.43759}{{\mathrm{SNR}}_{\max }+72.91711}-1)+11.53884;$

Step 200, set up sweet orange storage time Comprehensive Model: $\mathrm{Time}(F,{\mathrm{SNR}}_{\max })=P_1\&times;3.89102\ln \frac{F-4.00024}{3.70916}+{\mathrm{<figure-callout\; id="2"\; label="P"\; filenames="HDA0000370360320000032.png"\; state="\{\{state\}\}">P</figure-callout>}}_2\&times;\{1.58287\ln (\frac{1.43759}{{\mathrm{SNR}}_{\max }+72.91711}-1)+11.53884\}$ , P ₁And P ₂For undetermined coefficient;

By Time=1, Time ₁The signal to noise ratio (S/N ratio) eigenwert F=10.2 of=1 o'clock, Time ₂The SNR of=1 o'clock _max=-72.17; And Time=2, Time ₁The signal to noise ratio (S/N ratio) eigenwert F=11 of=2 o'clock, Time ₂The SNR of=2 o'clock _max=-72.34 substitution sweet orange storage time Comprehensive Models, thus P obtained ₁=-5, P ₂=4.467;

Thereby obtain sweet orange storage time Comprehensive Model:

Step 300, utilize step a and b and steps d to h, to detect the sweet orange sample W stored under 4 ℃ of environment, obtains signal to noise ratio (S/N ratio) eigenwert F=20 and the signal to noise ratio (S/N ratio) maximal value SNR of sweet orange sample W _max=-72.86; By SNR _maxIn frequency F substitution sweet orange storage time Comprehensive Model, calculate the storage time Time of sweet orange sample W _Prediction=6 days.

Use gas-detecting device and acoustic surface wave detection device to detect 20 the known sweet orange samples of storage time that store under 4 ℃ of environment, obtain 20 signal to noise ratio (S/N ratio) eigenwert F and 20 signal to noise ratio (S/N ratio) maximal value ANR _max, in substitution formula sweet orange storage time Comprehensive Model, obtain 20 storage time predicted value Time _Prediction, utilize formula

Calculate respectively predicated error, and calculate the mean value of predicated error, obtaining predicated error mean value is 0.93683, shows and utilizes gas-detecting device and acoustic surface wave detection device to predict more accurately the sweet orange storage time.

The present invention can quick nondestructive detects sweet orange fruit, and what surface acoustic wave detected reflection is the fruit internal structural information, is the smell information of fruit outside and gas detects reflection, and this combination can accurately characterize the situation of change in the sweet orange storage process.The method is conducive to instruct such fruit timely collecting, the decline of as far as possible avoiding fruits nutrition that harvest in advance causes to be worth, and also can reducing gathers evening causes the corrupt loss brought.

Should be understood that the present embodiment only is not used in and limits the scope of the invention be used to the present invention is described.Should be understood that in addition those skilled in the art can make various changes or modifications the present invention after the content of having read the present invention's instruction, these equivalent form of values fall within the application's appended claims limited range equally.

Claims (8)

1. the detection method in a sweet orange storage time, is characterized in that, comprises the steps:

(1-1) select n the sweet orange sample of just having plucked, the sweet orange sample be placed in congealer and store, the sweet orange sample is detected:

(1-1-1) setting the detection number of days is m ₁, the sequence number in storage time is Time ₁, Time ₁Initial value be 1; Use acoustic surface wave detection device to detect the frequency response of sweet orange, and obtain sweet orange storage time first predictor formula:

Described acoustic surface wave detection device comprises counter (1), and shielded box (2) is located at oscillator (3) and SAW (Surface Acoustic Wave) resonator (4) in shielded box; Oscillator and SAW (Surface Acoustic Wave) resonator form oscillation circuit, and counter is electrically connected to oscillation circuit, and counter is provided with for the wireless launcher (19) with computing machine (9) wireless connections, and SAW (Surface Acoustic Wave) resonator is provided with two electrodes;

Step a, optional 1 sweet orange sample in n sweet orange sample, and the sweet orange sample is put into to shielded box, two electrodes of SAW (Surface Acoustic Wave) resonator are contacted with the opposite flank of sweet orange (17) sample respectively;

Step b, acoustic surface wave detection device work is after 40 to 60 minutes, and counter gathers the frequency response curve of oscillation circuit, on frequency response curve, gathers 80 to 110 frequency values, and each frequency values is formed to frequency signal Input (t);

In computing machine, be preset with the first stochastic resonance system model

Wherein, ξ (t) is white Gaussian noise, and A, a and b are all constants, f ₀Be frequency modulating signal, D is noise intensity, and t is the Brownian movement Particles Moving time, and x is the coordinate of Particles Moving;

By in frequency signal Input (t) input the first stochastic resonance system model, the first stochastic resonance system model is resonated;

Computing machine utilizes formula $\mathrm{SNR}=2\left[\underset{\mathrm{\&Delta;\&omega;}\&RightArrow;0}{1\mathrm{im}}\underset{\mathrm{\&Omega;}-\mathrm{\&Delta;\&omega;}}{\overset{\mathrm{\&Omega;}+\mathrm{\&Delta;\&omega;}}{\&Integral;}}S\left(\mathrm{\&omega;}\right)\mathrm{d\&omega;}\right]/S_N\left(\mathrm{\&Omega;}\right)$ Calculate output signal-to-noise ratio SNR; Wherein, ω is signal frequency, and Ω is angular frequency, and S (ω) is signal spectral density, S _N(Ω) be the noise intensity in signal frequency range;

Computing machine draws the output signal-to-noise ratio curve in the first stochastic resonance system model, obtains two peak values of signal to noise ratio (S/N ratio) curve, using the absolute value of the difference of two peak values as signal to noise ratio (S/N ratio) eigenwert F; And signal to noise ratio (S/N ratio) eigenwert F is stored in computing machine;

Step c, work as Time ₁<m ₁, make Time ₁Value increases by 1, and will be detected the sweet orange sample and put back in congealer and store, after 24 hours, repeating step a and step b;

Obtain m ₁Individual and Time ₁The F be associated, by F and Time ₁Form point (F, Time ₁), according to m ₁Individual point (F, Time ₁) carry out linear fit, obtain sweet orange storage time first predictor formula: Time ₁=f (F);

(1-1-2) setting the detection number of days is m ₂, the sequence number in storage time is Time ₂, Time ₂Initial value be 1; Use gas-detecting device to detect the smell of sweet orange volatilization, and obtain sweet orange storage time second predictor formula:

Described gas-detecting device comprises air chamber (10), is located at sensor array (11) and analog to digital converter (12) in air chamber; Described air chamber is provided with sampling probe (13) and cleans probe (14), and sampling probe is provided with sampling air pump (15), and the cleaning probe is provided with and cleans air pump (16); Sensor array is electrically connected to analog to digital converter, is equipped with the data-interface be electrically connected to for computing machine on analog to digital converter, sampling air pump and cleaning air pump; Described sensor array comprises several gas sensors;

Steps d, computer control cleaning gas pump work, pure air passes in air chamber each gas sensor is cleaned, and when the response of each gas sensor is stablized to baseline, closes the cleaning air pump;

Step e, optional 1 sweet orange sample, inserted the sweet orange sample in sample bottle in n sweet orange sample, by sample bottle sealing and standing 30 to 40 minutes, sampling probe and air pressure balancer are injected in sample bottle simultaneously to the escaping gas that computer control sampling probe collected specimens produces; When gathering escaping gas, air pressure balancer will import in sample bottle through the air of activated carbon filtration, realizes air pressure balance;

Step f, escaping gas contacts with each sensor, and each sensor produces respectively analog response signal; Analog to digital converter is converted to respectively the digital response signal by each analog response signal, and computing machine is averaged digital response signal, obtains sensor array digital response signal I (t);

In computing machine, be preset with the second accidental resonance model

Wherein, ξ (t) is white Gaussian noise, and t is the Brownian movement Particles Moving time, and x is the coordinate of Particles Moving, f ₀Be signal frequency, a, b and A are constant;

Step g, by I (t) input the second accidental resonance model, make the second stochastic resonance system model produce accidental resonance;

Computing machine utilizes formula $\mathrm{SNR}=2\left[\underset{\mathrm{\&Delta;\&omega;}\&RightArrow;0}{1\mathrm{im}}\underset{\mathrm{\&Omega;}-\mathrm{\&Delta;\&omega;}}{\overset{\mathrm{\&Omega;}+\mathrm{\&Delta;\&omega;}}{\&Integral;}}S\left(\mathrm{\&omega;}\right)\mathrm{d\&omega;}\right]/S_N\left(\mathrm{\&Omega;}\right)$ Calculate output signal-to-noise ratio SNR; Wherein, ω is signal frequency, and Ω is angular frequency, and S (ω) is signal spectral density, S _N(Ω) be the noise intensity in signal frequency range;

Step h, computing machine draw the output signal-to-noise ratio curve of the second stochastic resonance system model, and the signal to noise ratio (S/N ratio) maximal value of choosing the signal to noise ratio (S/N ratio) curve in the signal to noise ratio (S/N ratio) curve is eigenwert SNR _max, and by SNR _maxBe stored in computing machine;

Step I, work as Time ₂<m ₂, make Time ₂Value increases by 1, will be detected the sweet orange sample and put back in congealer and store, and after 24 hours, repeating step d is to step h;

Obtain m ₂Individual and Time ₂The SNR be associated _max, by SNR _maxAnd Time ₂Form point (SNR _max, Time ₂), according to m point (SNR _max, Time ₂) carry out the Nonlinear Boltzmann match, obtain sweet orange storage time second predictor formula Time ₂=f (SNR _max);

(1-2) set up sweet orange storage time Comprehensive Model: Time (F, SNR _max)=P ₁* f (F)+P ₂* f (SNR _max), P ₁And P ₂For undetermined coefficient; Two groups of sweet orange storage times, the signal to noise ratio (S/N ratio) eigenwert F corresponding with the described storage time are reached to the SNR corresponding with the described storage time _maxSubstitution sweet orange storage time Comprehensive Model, thus P obtained ₁And P ₂

(1-3) detect with described sweet orange sample same stored condition under sweet orange sample W to be detected, obtain the signal to noise ratio (S/N ratio) maximal value SNR of sweet orange sample W _maxWith signal to noise ratio (S/N ratio) eigenwert F; By SNR _maxIn frequency F substitution sweet orange storage time Comprehensive Model, calculate the storage time Time of sweet orange sample W _Prediction.

2. the detection method in a kind of sweet orange storage time according to claim 1, is characterized in that, the centre frequency of described SAW (Surface Acoustic Wave) resonator is 433.92MHz.

3. the detection method in a kind of sweet orange storage time according to claim 1, is characterized in that, described n is 6 to 15.

4. the detection method in a kind of sweet orange storage time according to claim 1, is characterized in that, described m ₁And m ₂Be 10 to 180.

5. the detection method in a kind of sweet orange storage time according to claim 1, it is characterized in that, described gas sensor is 8, be respectively the first sensor for detection of sulfide gas, the second sensor for detection of inflammable gas, the 3rd sensor for detection of Ammonia gas, four-sensor for detection of ethanol class gas, the 5th sensor for detection of hydrocarbon component gas, the 6th sensor for detection of alkanes gas, for detection of the 7th sensor of propane and butane, for detection of the 8th sensor of nitride gas.

6. the detection method in a kind of sweet orange storage time according to claim 1, it is characterized in that, described SAW (Surface Acoustic Wave) resonator comprises piezoelectric substrate (5), interdigital transducer (6), 2 reflecting gratings (7) and is located at 4 gain grid (8) of reflecting grating both sides; Described electrode is electrically connected to the link (18) on being located at interdigital transducer.

7. according to claim 1 or 2 or 3 or 4 or the detection method in 5 or 6 described a kind of sweet orange storage times, it is characterized in that, in described step c, pure air passes in air chamber each gas sensor was cleaned 40 to 50 minutes.

8. according to claim 1 or 2 or 3 or 4 or the detection method in 5 or 6 described a kind of sweet orange storage times, it is characterized in that, in described step c, pure air cleans sensor with the flow velocity of 860mL/min to 1100mL/min.

Images