CN111579736B - Method for controlling processing production degree and evaluating quality of gardenia jasminoides ellis - Google Patents

Abstract

Aiming at the control method of using personal experience to judge the color as the only index in the processing and production process of the traditional Chinese medicine Gardenia, the present invention introduces a new method for processing gardenia based on smell. The electronic nose technology that can accurately quantify the odor value is used to collect the sensor response value that changes during the processing of Gardenia, effectively solving the technical problem of lack of objectivity in the evaluation process. The methods of digital standard, model standard and "heat" discriminant function are used for on-line control in the processing of gardenia decoction pieces. And by means of the correlation analysis between electronic nose and gas chromatography-mass spectrometry analysis, the indicator components that cause the change of the processed gardenia odor were determined, and the limit was established. Using this method in the processing and production of gardenia can ensure the stability of the quality of the decoction pieces of gardenia, and has the advantages of objectivity, rapidity and reliability, and solves the technical problem of lack of objectivity in human judgment in the prior art.

Description

A kind of method of gardenia processing production degree control and quality evaluation

technical field

The invention relates to the processing technology and quality control field of Chinese herbal decoction pieces of Gardenia, in particular to a discrimination method for controlling the processing degree and quality of decoction pieces of Gardenia by taking odor data as a new important index.

Background technique

Gardenia is the dried and ripe fruit of Gardenia jasminoides Eills, a Rubiaceae plant. In the clinical application of Gardenia, there are often different pieces of raw gardenia, fried gardenia, coke gardenia and gardenia charcoal. The descriptions of the processing degree and finished product properties of the processed products of different specifications of Gardenia in "Chinese Medicine Processing Science" are based on the appearance and color as an important indicator. The details are as follows: the surface of the raw product is red-yellow or brown-red, the inner surface is lighter, the seeds are dark red or red-yellow; the surface of fried gardenia is yellow-brown or yellow-red; the surface of burnt gardenia is burnt brown, burnt yellow or burnt black, and the peel The inner surface is brown, and the surface of the seeds is yellow-brown or tan; the surface of gardenia charcoal is dark brown or burnt black. It can be seen that at present gardenia only uses artificial evaluation color as its only indicator during processing, but in the actual production process, with the increase of processing temperature and time, gardenia will have obvious changes in smell, such as producing Obvious burnt aroma. Since the "smell" index is difficult to control in actual evaluation, it has never been included in the standard for testing the processing degree and quality of gardenia, but in fact, adding this index can ensure the processing control and quality standards of gardenia pieces. more reasonable and stable.

The change of the smell of traditional Chinese medicine is one of the means of judging the degree of processing of traditional Chinese medicine, that is, during the processing of traditional Chinese medicine, the degree of processing (heat) of traditional Chinese medicine is often described by the change of smell, but the key factor of the processing process of traditional Chinese medicine is "heat" so far. It is still a vague concept, and its scientific connotation has not yet been elucidated. At the same time, the smell of decoction pieces is also one of the important indicators for the quality evaluation of Chinese herbal medicine pieces, which has a long-term practical basis. However, the character indicators such as "smell" are relatively "fuzzy" in the description of the quality control of the indicators, such as "scent escape", "burnt aroma"... The evaluation method based on human sense of smell will be affected by factors such as the environment and the evaluator. The impact of decoction does not meet the objective requirements, that is, data-based measurement standards, which is not conducive to the implementation of the standardization of traditional Chinese medicine processing technology, the quality of decoction pieces cannot be stably controlled, and there is no judgment standard in the supervision of decoction pieces, making law enforcement actions difficult. Therefore, the objective quantification of the indicators of the appearance of decoction pieces represented by "smell" has become an important problem to be solved urgently in the Chinese medicine decoction pieces industry.

With the rapid development of modern bionic technology and the update of sensor technology, electronic nose technology that simulates human sense of smell can effectively solve the problem of lack of objectivity in odor evaluation. The electronic nose is an artificial olfactory system, which imitates the human odor recognition mechanism. The sensor array adsorbs odor molecules and generates signals, simulating the process of odor molecules binding to receptor proteins on the surface of human olfactory cells; the generated signals are processed by the signal processing system. And transmission, the simulation signal is further processed and amplified by the olfactory cell neural network and the olfactory glomerulus; finally, the pattern recognition system makes judgments on the processed signals, simulating the process of the human brain making judgments on odors.

The raw data curve detected by the electronic nose represents the change process of the response intensity of each sensor over time. When the volatile gas reaches the measurement chamber, the resistance value of the electronic nose sensor will change due to the redox reaction. The positive response value of the sensor represents that the effect of reducing gas is greater than that of oxidizing gas, while the negative response value represents the effect of oxidizing gas. Greater than the reducing gas, its relative resistance change rate is used as the response of the electronic nose sensor to the sample, which is also called the odor response intensity. The calculation process is "R=(R ₀ -R _t )/R ₀ ", where R is the sensor response value, R ₀ is the resistance value of the sensor at 0 seconds, and R _t is the instantaneous resistance value of the sensor when it touches the sample odor. During the detection process, the resistance value change of the sensor is recorded and calculated every 1 second. In addition, the maximum response intensity value of each sensor, that is, the peak or trough data of the curve, has a low relative standard deviation (RSD) when measuring the same sample, and is usually the most discriminative for different samples. Therefore, in the data analysis and processing of the electronic nose, most of the peaks or valleys in the original response curve of the sensor are selected as the characteristic points for analysis.

Aiming at the blank that there is no "smell" index in the quality control of decoction pieces of Gardenia, and the characteristics of "smell" being highly subjective and difficult to control, the present invention introduces a technology that can accurately quantify the smell of gardenia, namely the electronic nose technology for gardenia The method of measuring the smell of each decoction piece, quantifying its odor data, formulating its relevant evaluation standards, and realizing the identification and evaluation of each decoction piece of Gardenia; and combined with the electronic nose data to clarify the material composition that causes the change of the smell of Gardenia, as the evaluation of the quality of the decoction piece. Indicators The present invention is applied to the control in the processing process of the decoction pieces of gardenia, to ensure the objectivity, speed and reliability of the detection data, so as to ensure the stable quality of the decoction pieces of gardenia.

SUMMARY OF THE INVENTION

In order to solve the above-mentioned problems, the purpose of the present invention is to provide a method for the control and quality evaluation of processed gardenia based on odor value. In order to achieve this purpose, the following technical solutions are adopted:

In the method, the percentile method, discriminant function and DFA analysis are performed on the odor data obtained after the electronic nose detection of four different pieces of Gardenia jasminoides, and the correlation analysis between the GC-MS method and the electronic nose is used to determine the odor change caused by the method. Therefore, the method can effectively solve the problems that the degree of the processing of the gardenia is not judged effectively and the quality control of the decoction pieces is not in place. details as follows:

The invention provides a method for controlling and evaluating the quality of processed gardenia, comprising the following steps: measuring the gardenia odor value with an electronic nose; The scientific expression function of "heat" established by the Bayesian discriminant method, or the discriminant model established by the discriminant factor analysis, to control the degree of the processing process of the gardenia and the quality of the processed gardenia; or use the gas chromatography-mass spectrometry method to analyze the components The content and component content all meet the specified range formulated by the content limit method to control the degree of the processing process of gardenia and the quality of processed gardenia products.

Method 1: Use the electronic nose to measure the odor value of Gardenia, all the odor values conform to the specified range established by the percentile method, or conform to the scientific expression function of "heat" established by the Bayesian discriminant method, or conform to the discriminant factor analysis The established discriminant model to control the degree of the processing process and the quality of the processed gardenia products

The gardenia odor value is the sensor response value of the electronic nose, and its determination includes the following steps: taking a sample, setting the carrier gas type and flow rate, setting headspace incubation parameters, headspace injection parameters, acquisition parameters, and odor detection; Described gardenia processing production degree and the quality control of gardenia processing product comprise the following steps: use the response value of 18 or 8 sensors to all conform to the specified range established by the percentile method, or use the response value of 8 sensors In line with the scientific expression function of "heat" established by the Bayesian discriminant method, it is determined that the samples are processed into the corresponding types of decoction pieces.

Further, the gardenia odor value is the sensor response value of the electronic nose, and the measurement includes the following steps: taking a sample, setting the carrier gas type and flow rate, setting headspace incubation parameters, headspace injection parameters, collection parameters, and odor Detecting; the processing production degree of the gardenia and the quality control of the processed gardenia product include the following steps: using the discriminant model established by the discriminant factor analysis, the sensor response value is subjected to discriminant factor analysis, and the three characteristic factors with the largest contribution rate are used to obtain a three-dimensional For the score map, for the unknown sample, after the response values of the 8 sensors obtained by the electronic nose detection, the corresponding decoction piece types are determined according to the degree of coincidence with the corresponding area of the three-dimensional map. The headspace incubation parameters include incubation time, incubation temperature, and agitation speed; the headspace injection parameters include injection volume, injection speed, and injection needle temperature; and the acquisition parameters include acquisition time, flushing time, and lag time.

Further, the number of the sensors is 18, and the specified range of the response value is as follows:

The odor value range of raw gardenia also conforms to: the odor value of sensor LY2/LG is 0.041～0.157, the odor value of sensor LY2/G is -0.231～-0.101, the odor value of sensor LY2/AA is -0.164～-0.077, the odor value of sensor LY2/ The smell value of GH is -0.316～-0.129, the smell value of sensor LY2/gCTl is -0.251～-0.101, the smell value of sensor LY2/gCT is -0.064～-0.029, the smell value of sensor T30/1 is 0.394～0.610, the smell value of sensor P10/ The smell value of sensor 1 is 0.566～0.731, the smell value of sensor P10/2 is 0.442～0.526, the smell value of sensor P40/1 is 0.511～0.624, the smell value of sensor T70/2 is 0.425～0.651, and the smell value of sensor PA/2 is 0.469～0.469～ 0.656, the smell value of sensor P30/1 is 0.599～0.760, the smell value of sensor P40/2 is 0.596～0.733, the smell value of sensor P30/2 is 0.663～0.839, the smell value of sensor T40/2 is 0.308～0.421, the smell value of sensor T40/1 The odor value of sensor TA/2 is 0.337～0.366, and the odor value of sensor TA/2 is 0.392～0.484.

The odor value range of fried gardenia also conforms to: the odor value of sensor LY2/LG is 0.028～0.496, the odor value of sensor LY2/G is -0.484～-0.086, the odor value of sensor LY2/AA is -0.358～-0.066, the odor value of sensor LY2/ The smell value of GH is -0.717～-0.112, the smell value of sensor LY2/gCTl is -0.640～-0.084, the smell value of sensor LY2/gCT is -0.133～-0.026, the smell value of sensor T30/1 is 0.355～0.824, the smell value of sensor P10/ The smell value of sensor 1 is 0.539～0.875, the smell value of sensor P10/2 is 0.433～0.646, the smell value of sensor P40/1 is 0.495～0.765, the smell value of sensor T70/2 is 0.383～0.854, and the smell value of sensor PA/2 is 0.443～0.443～ 0.850, the smell value of sensor P30/1 is 0.573～0.906, the smell value of sensor P40/2 is 0.571～0.877, the smell value of sensor P30/2 is 0.616～0.939, the smell value of sensor T40/2 is 0.288～0.567, the smell value of sensor T40/1 The odor value of sensor TA/2 is 0.338～0.495, and the odor value of sensor TA/2 is 0.387～0.667.

The odor value range of Jiaojiazi also conforms to: the odor value of sensor LY2/LG is 0.049～0.295, the odor value of sensor LY2/G is -0.345～-0.118, the odor value of sensor LY2/AA is -0.241～-0.089, the odor value of sensor LY2/ The smell value of GH is -0.473～-0.153, the smell value of sensor LY2/gCTl is -0.389～-0.120, the smell value of sensor LY2/gCT is -0.058～-0.029, the smell value of sensor T30/1 is 0.373～0.596, the smell value of sensor P10/ The odor value of sensor 1 is 0.559～0.726, the odor value of sensor P10/2 is 0.439～0.523, the odor value of sensor P40/1 is 0.503～0.617, the odor value of sensor T70/2 is 0.417～0.646, and the odor value of sensor PA/2 is 0.473～0.473～ 0.657, the smell value of sensor P30/1 is 0.586～0.777, the smell value of sensor P40/2 is 0.594～0.723, the smell value of sensor P30/2 is 0.648～0.837, the smell value of sensor T40/2 is 0.297～0.418, the smell value of sensor T40/1 The odor value of sensor TA/2 is 0.331～0.359, and the odor value of sensor TA/2 is 0.387～0.481.

The odor value range of gardenia charcoal also conforms to: the odor value of sensor LY2/LG is 0.036～0.149, the odor value of sensor LY2/G is -0.177～-0.079, the odor value of sensor LY2/AA is -0.132～-0.060, the odor value of sensor LY2/ The smell value of GH is -0.223～-0.101, the smell value of sensor LY2/gCTl is -0.178～-0.076, the smell value of sensor LY2/gCT is -0.045～-0.019, the smell value of sensor T30/1 is 0.298～0.523, the smell value of sensor P10/ The smell value of sensor 1 is 0.492～0.663, the smell value of sensor P10/2 is 0.414～0.483, the smell value of sensor P40/1 is 0.465～0.568, the smell value of sensor T70/2 is 0.323～0.559, and the smell value of sensor PA/2 is 0.396～0.396～ 0.593, the smell value of sensor P30/1 is 0.491～0.709, the smell value of sensor P40/2 is 0.527～0.675, the smell value of sensor P30/2 is 0.535～0.776, the smell value of sensor T40/2 is 0.251～0.376, the smell value of sensor T40/1 The odor value of sensor TA/2 is 0.326～0.342, and the odor value of sensor TA/2 is 0.369～0.427.

Further, there are 8 sensors, which are determined by the specified range of sensor response values or the scientific expression function of "heat":

Judgment method 1: Use the specified range of sensor response value to judge:

The odor value range of raw gardenia also conforms to: the odor value of sensor LY2/LG is 0.041～0.157, the odor value of sensor LY2/G is -0.231～-0.101, the odor value of sensor LY2/gCTl is -0.251～-0.101, the odor value of sensor T30/ The smell value of sensor 1 is 0.394～0.610, the smell value of sensor PA/2 is 0.469～0.656, the smell value of sensor P30/1 is 0.599～0.760, the smell value of sensor T40/1 is 0.337～0.366, and the smell value of sensor TA/2 is 0.392～0.392～ 0.484.

The odor value range of fried gardenia also conforms to: sensor LY2/LG odor value 0.028～0.496, sensor LY2/G odor value -0.484～-0.086, sensor LY2/gCTl odor value -0.640～-0.084, sensor T30/ The smell value of sensor 1 is 0.355~0.824, the smell value of sensor PA/2 is 0.443~0.850, the smell value of sensor P30/1 is 0.573~0.906, the smell value of sensor T40/1 is 0.338~0.495, and the smell value of sensor TA/2 is 0.387~ 0.667.

The odor value range of Jiaojiazi also conforms to: the odor value of sensor LY2/LG is 0.049～0.295, the odor value of sensor LY2/G is -0.345～-0.118, the odor value of sensor LY2/gCTl is -0.389～-0.120, the odor value of sensor T30/ The smell value of sensor 1 is 0.373～0.596, the smell value of sensor PA/2 is 0.473～0.657, the smell value of sensor P30/1 is 0.586～0.777, the smell value of sensor T40/1 is 0.331～0.359, and the smell value of sensor TA/2 is 0.387～ 0.481.

The odor value range of gardenia charcoal also conforms to: the odor value of sensor LY2/LG is 0.036～0.149, the odor value of sensor LY2/G is -0.177～-0.079, the odor value of sensor LY2/gCTl is -0.178～-0.076, the odor value of sensor T30/ The smell value of sensor 1 is 0.298～0.523, the smell value of sensor PA/2 is 0.396～0.593, the smell value of sensor P30/1 is 0.491～0.709, the smell value of sensor T40/1 is 0.326～0.342, and the smell value of sensor TA/2 is 0.369～ 0.427.

Or judgment method 2: use the scientific expression function of "heat" to judge:

Raw gardenia "heat" function:

F ₁ =175.9SR _LY2/LG -12701.9SR _LY2/G +13775.6SR _LY2/gCTl -9777.4SR _T30/1 -2684.5SR _PA/2 +9908.0SR _P30/1 +3873.3SR _T40/1 +15272.5SR _TA/2 -4007.3

Fried gardenia "heat" function:

F ₂ =301.7SR _LY2/LG -12571.5SR _LY2/G +13779.1SR _LY2/gCTl -10659.4SR _T30/1 -1898.8SR _PA/2 +10019.2SR _P30/ 1+3539.7SR _T40/1 +15667.0SR _TA/2 -4131.4

Jiao Gardenia "heat" function:

F ₃ =367.7SR _LY2/LG -12050.5SR _LY2/G +13224.5SR _LY2/gCTl -11336.2SR _T30/1 -613.4SR _PA/2 +9634.9SR _P30/1 +3192.1SR _T40/1 +15123.8SR _TA/2 -3930.5

Gardenia charcoal "heat" function:

F ₄ =538.1SR _LY2/LG -12128.8SR _LY2/G +13518.9SR _LY2/gCTl -10888.0SR _T30/1 -908.6SR _PA/2 +9472.6SR _P30/1 +3385.9SR _T40/1 +15052.4SR _TA/2 -3885.9

Among them, SR is the response value of the sample on each sensor, and the subscript is the corresponding sensor name.

Or judgment method 3: Use discriminant factor analysis to establish a discriminant model, conduct discriminant factor analysis on the sensor response value, and use the three characteristic factors with the largest contribution rate to obtain a three-dimensional score map. Its coincidence degree with the corresponding area of the three-dimensional map is determined and processed into the corresponding type of decoction piece. The following eight sensors were used: LY2/LG, LY2/G, LY2/gCTl, T30/1, PA/2, P30/1, T40/1, TA/2.

The following is a further improvement of the present invention, the response value of the electronic nose sensor is the maximum response value as the calculated value, and the average value calculated after each sample is measured twice as the sample odor value.

The sample needs to be pulverized and passed through a No. 3 sieve, and 0.4 g of the sample is weighed for sample analysis; in the carrier gas parameters, the type of carrier gas is synthetic dry air with a flow rate of 150 mL/min; in the headspace incubation parameters, the incubation time is 300 s, The incubation temperature was 50°C, and the stirring speed was 500rpm; in the headspace injection parameters, the injection volume was 1500 μL, the injection speed was 500 μL/s, and the injection needle temperature was 60°C; in the acquisition parameters, the acquisition time was 120s, the flushing time was 120s, and the delay time was 600s.

Method 2: Use gas chromatography-mass spectrometry to analyze the content of the components, and the content of the components is all within the specified range formulated by the content limit method to control the degree of the processing process of gardenia and the quality of processed gardenia products

The content of the components was analyzed by gas chromatography-mass spectrometry, and the production degree of the processed gardenia and the quality of the processed gardenia products were controlled according to the content of the components all within the specified range. The specified range was methyl acetate, 2,5-dimethylpyridine The relative percentage content of oxazine, acetic acid, furfural, and 4-methylene isophorone is the evaluation index for the processing of gardenia: methyl acetate takes the relative percentage content in raw gardenia as 1, and calculates fried gardenia, coke gardenia The ratio of the relative percentage content of Zizi, gardenia charcoal and raw gardenia, fried gardenia, coke gardenia, gardenia charcoal should be controlled at 1.75-2.47, 1.74-2.71, 2.10-3.70; 2,5-dimethyl Pyrazine takes the relative percentage content of raw gardenia as 1, and calculates the ratio of the relative percentage content of fried gardenia, coke gardenia, gardenia charcoal and raw gardenia, fried gardenia, coke gardenia, and gardenia charcoal should be respectively Controlled at 9.76-15.64, 4.19-8.15, 1.86-3.30; the relative percentage of acetic acid in raw gardenia is 1, and the ratio of the relative percentage of fried gardenia, coke gardenia, gardenia charcoal and raw gardenia is calculated, Fried gardenia, coke gardenia and gardenia charcoal should be controlled at 5.13-7.40, 3.89-5.34, 2.48-3.48 respectively; the relative percentage of furfural in raw gardenia is 1, and the calculation of fried gardenia, coke gardenia and gardenia The ratio of the relative percentage content of seed charcoal to raw gardenia, fried gardenia, coke gardenia, gardenia charcoal should be controlled at 246.60-397.65, 111.60-184.80, 61.20-101.75; The relative percentage content of raw gardenia is 1. Calculate the ratio of the relative percentage of fried gardenia, coke gardenia, gardenia charcoal to raw gardenia, and stir-fried gardenia, coke gardenia, and gardenia charcoal should be controlled at 12.11 respectively. -18.21, 14.11-18.81, 6.86-12.46.

beneficial effect

The method of monitoring the change of smell in the processing process introduced by the present invention to realize the control of the processing production degree and the quality evaluation of gardenia, aims to use an electronic nose to measure the smell of the gardenia sample, quantify its smell data, and combine with the GC-MS method, After statistical calculation, the gardenia odor characterized by digitization, modeling, function and limited quantity of specific substances is realized. Using this method in the on-line production of gardenia can ensure the stability of the quality of the decoction pieces of gardenia, and has the advantages of being objective, fast and reliable.

Description of drawings

Figure 1 is a graph showing the effect of different detection conditions of Gardenia on the RSD value of the maximum response value of the sensor. A: Pulverized particle size, B: Weighing amount, C: Injection amount, D: Incubation temperature, E: Incubation time

Fig. 2 is the linear discriminant analysis diagram of the sensor array of Gardenia decoction pieces. A: original array, B: optimized array; 1: raw gardenia, 2: fried gardenia, 3: coke gardenia, 4: gardenia charcoal

Figure 3 is a diagram of the DFA discriminant model for decoction pieces of Gardenia. SZZ: raw gardenia CZZ: fried gardenia JZZ: coke gardenia ZZT: gardenia charcoal

Figure 4 is a GC-MS image of the decoction pieces of Gardenia. A: GC-MS overlay, B: common patterns of GC-MS fingerprints (peaks 1, 3, 4, 5, 7, 10, 12, 20, 22, 24, 26, 33, 41, 68 , 78, 83, 106, 108, 110, 112, 121, 126, 132, 137, 140, 150, 178, 222, 236, 278 are the 30 common chromatographic peaks of gardenia samples)

Figure 5 is a graph showing the change trend of the relative percentage content of various compounds during the frying process of Gardenia jasminoides.

Detailed ways

In conjunction with specific embodiment, the present invention is further described as follows:

Embodiment 1 Determination of Gardenia Pieces Odor Detection Method

During the experiment, a batch of decoction pieces of Gardenia was selected as the experimental object for single-factor investigation. FOX-4000 electronic nose with Alpha Soft 11.0 software (Alpha MOS, France) was used as the odor detection instrument. The absolute value of the response value of the sample on the electronic nose sensor is controlled within the range of 0.3-0.9 and the response value of the sample has good stability (the RSD value of the response value is small) as the investigation indicators. The experimental parameters were optimized in terms of crushing particle size, sample injection volume, incubation temperature and incubation time of the instrument.

1.1 Optimization of detection parameters

1.1.1 Optimization of crushing particle size

Take the original decoction pieces of Gardenia, pass through the No. 1 sieve, No. 2 sieve, No. 3 sieve, and No. 4 sieve, a total of five samples, weigh 0.4g of the samples respectively, put them into the headspace sampling bottle, control the injection volume to 1500μL, incubate The detection was carried out under the same conditions as the temperature of 50 °C and the incubation time of 300 s, with 6 replicates for each level. The results showed that the maximum response value of each sensor gradually increased with the increase of the pulverization degree of the decoction pieces. When the decoction piece powder passed through the No. 4 sieve, the odor response value reached the maximum. Combined with the RSD change of the maximum response value of each sensor, it was found that the RSD of the response value of each sensor was the lowest and the most stable when the decoction powder passed through the No. 3 sieve. Taking comprehensive consideration to determine the degree of pulverization of the decoction pieces, the powder passes through the No. 3 sieve. The RSD value results for the maximum response value are shown in Figure 1(A).

1.1.2 Optimization of weighing sample

Weigh 0.1g, 0.2g, 0.3g, 0.4g, 0.5g of gardenia sample powder (passed through No. 3 sieve) respectively, put them into headspace injection bottles, control the injection volume to 1500μL, incubation temperature of 50°C and incubation time 300s and other conditions were tested under the same conditions, and each level was replicated 3 times. The results show that the maximum response value of each sensor increases gradually with the increase of the weighing sample. Combined with the RSD change of the maximum response value of each sensor, it is found that the RSD of the response value of each sensor is the lowest and the most stable when the sample weight is 0.4g. Taking comprehensive consideration, the weighing sample size is set at 0.4g. The RSD value results for the maximum response value are shown in Figure 1(B).

1.1.3 Optimization of injection volume

Weigh 0.4 g of gardenia sample powder (passed through a No. 3 sieve), put it into a headspace injection bottle, control the incubation temperature of 50 °C and the incubation time of 300 s under the same conditions, set the injection volume to 500 μL, 1000 μL, 1500 μL and 2000 μL in 3 replicates for each level. The results showed that the maximum response value of each sensor gradually increased with the increase of the injection volume. At the same time, when the injection volume was 1500 μL, the RSD of the maximum response value of each sensor was the lowest and more stable. The injection volume was determined to be 1500 μL after comprehensive consideration. The RSD value results for the maximum response value are shown in Figure 1(C).

1.1.4 Optimization of incubation temperature

Weigh 0.4 g of gardenia sample powder (passed through a No. 3 sieve), put it into a headspace injection bottle, control the injection volume to 1500 μL, and set the incubation temperature to be 40 °C, 45 °C, and 300 s under the same conditions. °C, 50 °C, 55 °C, 6 copies for each level. The results showed that the maximum response value of each sensor gradually increased with the rise of incubation temperature, and the maximum value appeared when the incubation temperature was 50 °C, and the response value at 55 °C was close to it. At the same time, the maximum response value of each sensor was at 50 °C has the lowest and most stable RSD. The incubation temperature was determined to be 50°C after comprehensive consideration. The RSD value results for the maximum response value are shown in Figure 1(D).

1.1.5 Optimization of incubation time

Weigh 0.4 g of gardenia sample powder (passed through a No. 3 sieve), put it into a headspace injection bottle, control the injection volume to 1500 μL, and set the incubation time to 150s, 300s, and 600s under the same conditions as the incubation temperature of 50 °C. , 6 copies of each level, observe the change of the maximum response value of the sensor and the RSD value of the maximum response value to optimize the incubation time. The results showed that the maximum response value of each sensor gradually increased with the increase of incubation time, and the maximum value appeared when the incubation time was 300 s, and then the response value of individual sensors decreased slightly. Combined with the RSD change of the maximum response value of each sensor, when the incubation time was 300s, the RSD of the maximum response value of each sensor was the lowest and the most stable. Taking comprehensive consideration, the incubation time was set as 300s. The RSD value results for the maximum response value are shown in Figure 1(E).

According to the above results, the optimal conditions for gardenia electronic nose detection are determined as shown in Table 1.

Table 1 Gardenia electronic nose detection parameters

1.2 Methodological investigation

According to the optimal conditions determined by the optimization results, a methodological investigation was carried out.

1.2.1 Repeated investigation

Take 6 samples of the same batch of powder, put them into headspace injection bottles and seal them respectively, and use the best detection conditions for measurement. The results show that the RSD value of the 18 sensors is 3.799% at most, indicating good repeatability. The specific results are shown in Table 2.

Table 2 The results of repeatability investigation of gardenia electronic nose detection method

1.2.2 Stability investigation

Take the same batch of sample powder, put it into the headspace sampling bottle and seal it, and use the best detection conditions to measure at 0, 2, 4, 6, 8, and 10 hours, respectively. The RSD value of the maximum response value of the sensor is at most 4.908%, indicating that the sample is stable within 10 hours, and the results are shown in Table 3.

Table 3 The results of stability investigation of gardenia electronic nose detection method

1.3 Sample odor detection

A total of 121 batches of samples of decoction pieces of Gardenia (57 batches of raw gardenia, 32 batches of fried gardenia, 17 batches of coke gardenia, and 15 batches of gardenia charcoal) were tested by using the best detection conditions of electronic nose. Each batch of samples was repeated twice. 242 sets of data are obtained.

Example 2 Optimization of electronic nose sensor

2.1 Sensor Optimization

Since each sensor of the electronic nose has a broad-spectrum response to all gases, there is a high correlation between the maximum response intensities obtained by each sensor. This correlation results in a large amount of data redundancy to a certain extent. At the same time, some sensors are not sensitive to sample odor information, and their response is close to that of air. In order to simplify the sensor array and realize data dimensionality reduction, while eliminating redundant information, the integrity, validity and reliability of the acquired information are ensured, thereby reducing the complexity of pattern recognition, better realizing classification and discrimination, and optimizing the sensor.

The experiment adopts Wilks' Lambda method for stepwise discriminant analysis to optimize the electronic nose sensor. The F value was used as the discriminant statistic. Whether a variable can enter the model mainly depends on the significance level of the F test in the analysis of covariance and the set F value entering and leaving the model. The specific parameters are set to (default value): when F≥3.84, the variable enters the model; when F≤2.71, the variable moves out of the model. Stepwise discriminant analysis was carried out on the maximum response values of 18 sensors of gardenia samples, and the specific results are shown in Table 4.

Table 4 Stepwise discriminant analysis results of the maximum response value of 18 sensors of gardenia samples

As shown in Table 4, except the sensor LY2/gCT was kicked out of the model because the F value was less than 3.84 in step 8 of stepwise discrimination, LY2/LG, LY2/G, LY2/gCTl, T30/1, PA/2, P30 The statistics (F) of /1, T40/1 and TA/2 were all above 3.84, and the P values of these 8 variables were all less than 0.01. Finally, after 10 steps, the external and internal variables of the model have no input and no output, and the independent variable selection of the step-by-step discriminant analysis is completed. /1, PA/2, P30/1, T40/1, TA/2.

2.2 Sensor Optimization Verification

In order to compare the classification effects of different processed Gardenia products before and after the sensor array optimization, Linear Discriminant Analysis (LDA) was used to output the intuitive classification results. Linear discriminant analysis was used to analyze the original sensor array of all samples of Gardenia and the maximum response value of the optimized sensor array. The LDA analysis comparison before and after sensor optimization is shown in Figure 2. As shown in the figure, the total contribution rate of the discriminant graphs LD1 and LD2 of the sensor array optimized by Gardenia is 93.6%, indicating that the established linear discriminant function can explain most of the information and ensure the integrity and reliability of the obtained information. At the same time, the classification effect of the samples is significantly improved on the optimized sensor array, indicating that the optimized array eliminates redundant information to a certain extent, improves the data processing efficiency, and the optimized sensor array can replace the original array. Complete the task of identifying different pieces of Gardenia.

The establishment of embodiment 3 gardenia odor standard

In this experiment, the electronic nose odor response data of Gardenia decoction pieces (raw, fried, coke, and charcoal) were used, and based on the optimized sensor array, the digital standard and model standard of the sample odor were established respectively, which was the quality of the smell. Control provides the basis. The digital standard can express the odor standard in a digital form, which is beneficial to be directly included in the existing paper standard for evaluation and implementation, while the model standard expresses the odor standard in the form of a database, which can be adjusted and improved at any time with the increase of samples. This standard can be directly applied to the self-inspection process of enterprise samples.

3.1 Establishment of the digital standard for the smell of decoction pieces

In this experiment, based on the maximum response value of the optimized sensor array, the digital standard range of odor was established for the collected electronic nose odor data of each decoction piece. SPSS 23.0 software was used to analyze the collected odor data of different pieces of Gardenia. First, exploratory analysis was used to conduct multivariate normality test on the data of different sensors. The results showed that most of the original odor data of different pieces of Gardenia did not conform to the normality. Therefore, the P5 and P95 percentile indicators of the percentile method (Percentiles) were used to establish the odor number range of the bilateral 90% interval of each sensor of different decoction pieces of Gardenia. The results are shown in Tables 5 and 6.

Table 5 The odor digital standard of different pieces of Gardenia based on the original sensor array

Table 6 Numerical standards for the scent of different pieces of Gardenia based on the optimized sensor array

In order to judge the rationality of the reference range of the odor value of the decoction pieces, the independent sample test under the nonparametric test was used to statistically analyze the response values of each sensor of each decoction piece of Gardenia collected. The specific rank sum test results are shown in Table 7.

Table 7 Nonparametric test results of gardenia

As shown in Table 7, the response values of the sensors of the electronic nose in the four groups of different pieces of gardenia, fried, coke, and charcoal, were all less than 0.01, indicating that the odor figures of the four groups of different pieces of gardenia The differences between the ranges were all statistically significant, and the established odor digital standard range could better distinguish the decoction pieces of Gardenia with different specifications.

3.2 Establishment of the standard for the odor pattern of decoction pieces

Based on the optimized array, discriminant factor analysis (DFA) was used to establish discriminant models for the electronic nose odor data of Gardenia decoction pieces, and the models were evaluated by cross-validation score. The established DFA discriminant models and validation scores of different pieces of Gardenia are shown in Figure 3.

As shown in Figure 3, the total contribution rate of the discriminant factors DF1, DF2 and DF3 reaches 100%, which better reflects the original data information, and the processing results are reliable. DFA can distinguish different processed samples of Gardenia. Fried products are distributed in the negative direction of DF1, while coke and charcoal products are distributed in the positive direction of DF1; meanwhile, raw and fried products are distributed on both sides of DF2, and coke and charcoal products are distributed on both sides of DF3. The cross-validation score of the model was greater than 85, indicating that the established discriminant model for decoction pieces of Gardenia was better.

The DFA discriminant model for decoction pieces established above can intuitively discriminate unknown samples in a visual way. When an unknown sample enters the library created by the DFA discriminant model, when the sample is projected to the corresponding area, its attribution can be confirmed; otherwise, the unknown sample will be judged as "unrecognized". The established discriminant library will expand with the input of standard decoction pieces and is adjustable.

Embodiment 4. Establishment of the mathematical discrimination formula of "heat" processed by gardenia

Heat is a concept to describe the degree of processing of traditional Chinese medicine, but so far it is still at the level of experience and there is no scientific explanation. This experiment initially clarifies the scientific connotation of "heat" through the established mathematical discriminant function.

4.1 Establishment of the mathematical discrimination formula for the processing degree of decoction pieces

Using SPSS 23.0 (IBM Corporation, USA) software, the maximum response value of the sensor array optimized for Gardenia decoction pieces was analyzed by Bayesian discriminant, and the mathematical discriminant function formula of the processing temperature of Gardenia decoction pieces represented by sensor data was established. The coefficients of the respective variables of the function See Table 8.

Table 8 Gardenia Mathematical Discriminant Function Results

According to the coefficient of the heat discriminant function of the decoction pieces of Gardenia in Table 8, the Bayesian discriminant function of different processing degrees (heat) of Gardenia is established as follows:

Raw gardenia "heat" function:

F ₁ =175.9SR _LY2/LG -12701.9SR _LY2/G +13775.6SR _LY2/gCTl -9777.4SR _T30/1 -2684.5SR _PA/2 +9908.0SR _P30/1 +3873.3SR _T40/1 +15272.5SR _TA/2 -4007.3

Fried gardenia "heat" function:

F ₂ =301.7SR _LY2/LG -12571.5SR _LY2/G +13779.1SR _LY2/gCTl -10659.4SR _T30/1 -1898.8SR _PA/2 +10019.2SR _P30/ 1+3539.7SR _T40/1 +15667.0SR _TA/2 -4131.4

Jiao Gardenia "heat" function:

F ₃ =367.7SR _LY2/LG -12050.5SR _LY2/G +13224.5SR _LY2/gCTl -11336.2SR _T30/1 -613.4SR _PA/2 +9634.9SR _P30/1 +3192.1SR _T40/1 +15123.8SR _TA/2 -3930.5

Gardenia charcoal "heat" function:

F ₄ =538.1SR _LY2/LG -12128.8SR _LY2/G +13518.9SR _LY2/gCTl -10888.0SR _T30/1 -908.6SR _PA/2 +9472.6SR _P30/1 +3385.9SR _T40/1 +15052.4SR _TA/2 -3885.9

Note: SR is the response value of the sample on each sensor.

At the same time, this function can also be used as a method for the identification of decoction pieces of Gardenia: the gas value of each unknown decoction piece is substituted into the above four mathematical identification functions to calculate and compare the values of F1, F2, F3 and F4, and the highest value of F1 is determined as For raw gardenia, the highest F2 value is judged as fried gardenia, the highest F3 value is judged as coke gardenia, and the highest F4 value is judged as gardenia charcoal.

4.2 Interactive verification results

In order to judge the rationality of the discriminant function formula, the specific odor data of each sample of gardenia raw products, fried products, coke products and charcoal products with determined specifications were substituted into the established discriminant function formula, and then the ordinary method was adopted (Original) And cross-validated (Cross-validated) method to verify the discriminant correct rate of the established discriminant function, the results are shown in Table 9.

Table 9 Gardenia discriminant analysis verification results

Note: 1, 2, 3, and 4 in the vertical column represent the original attribution of gardenia samples, namely raw, fried, coke, and charcoal products; 1, 2, 3, and 4 in the horizontal column represent gardenia, respectively. The odor data of the sample is determined by the discriminant function calculation, that is, the raw product, the fried product, the coke product, and the charcoal product.

As shown in the table, for each record in 4 groups of gardenia samples (raw product, fried product, coke product, charcoal product), the accuracy rates of common verification methods are 99.1%, 96.9%, 82.4%, and 100.0%, respectively. ; The accuracy rate of the interactive verification method is 99.1%, 93.8%, 79.4% and 100% respectively. The misjudged samples mainly occurred in gardenia coke products, and a small number of coke products were judged as charcoal products, which may be due to the misjudgment caused by the proximity of coke products and charcoal products. The results show that the established discriminant function for predicting the processing degree of Gardenia has very few misjudgments of the samples, so the established discriminant function for the processing temperature of Gardenia decoction pieces is stable and reasonable.

Example 5 Determination of odor characteristic substances of decoction pieces of gardenia

The raw gardenia decoction pieces were processed according to the standard operating procedures for processing, and samples at different processing time points were respectively taken for odor detection. According to the detection results, 16 batches of samples with obvious odor changes were selected for GC-MS fingerprint analysis, and the preparation time was established. The matching chromatographic peak data of the decoction pieces is analyzed; the correlation degree between the multi-component matching data in the decoction pieces at different processing time points and the corresponding decoction piece odor data is analyzed, and the matching chromatographic peaks that are correlated with the response value changes of the electronic nose sensor are found to determine the cause. The component or component group that changes the smell during the processing of the decoction piece.

5.1 GC-MS Analysis of Gardenia Pieces

5.1.1 Instruments

Agilent 7890B-7000C GC-MS combined instrument with Agilent 7697A headspace autosampler.

5.1.2 Methods

Preparation of the test product: The decoction pieces of Gardenia were crushed and passed through a No. 3 sieve during the processing, and 1.5g of each decoction piece powder was accurately weighed and placed in a 10mL headspace sample bottle.

GC-MS conditions for Gardenia:

Headspace conditions: Equilibrium temperature 90°C, loop 105°C, transfer line 120°C, equilibration time 15min, carrier gas He, oscillating frequency 250 times/min, filling pressure 15psi, pressurization time 0.1min, injection The time is 0.5min.

Chromatographic conditions: HP-INNOWAX chromatographic column (30m×250μm, 0.25μm); inlet temperature 220°C; programmed temperature: initial temperature 45°C, 3°C/min to 110°C, hold for 10min, then 5°C /min to 150 °C, hold for 5 min, and then increase to 260 °C at 20 °C/min, the total analysis time is 50 min; the carrier gas is He, the constant flow mode, the flow rate is 1.0 ml/min; the split ratio is 5:1; Quantity 1mL.

Mass spectrometry conditions: EI ion source, electron energy 70 eV, ion source temperature 230 ℃; quadrupole temperature 150 ℃, interface temperature 270 ℃; scanning range 35-500 amu; electron multiplier voltage 1557.3 V; solvent delay 1 3.15min.

5.1.3 Results

The decoction pieces of Gardenia in the frying process were detected by GC-MS according to the above conditions. The GC-MS overlay of the sample is shown in Figure 4(A), and the GC-MS data was imported into the "Chinese Medicine Chromatographic Fingerprint Similarity Evaluation Software" (2012 edition of the National Pharmacopoeia Commission), the common pattern of the fingerprints of the decoction pieces generated is shown in Figure 4(B). After comparison, a total of 333 chromatographic peaks were detected in the Gardenia group, of which there were 30 common chromatographic peaks.

5.2 Correlation analysis and component confirmation of GC-MS and electronic nose data

Establish a data matrix of gardenia matching chromatographic peaks and the number of processed samples, and a data matrix of the sensor response value and the number of processed samples, and carry out gray correlation analysis between the two. The gray correlation degree r>0.9 is the inclusion standard, and it is determined that 52 These gas chromatographic peaks have a strong correlation with the electronic nose data of Gardenia decoction pieces.

By checking the standard chromatogram in the online search workstation database NIST 14.0 and referring to the literature for qualitative identification, a total of 36 compounds were identified in the 52 chromatographic peaks with high correlation. The specific results are shown in Table 10, including 1 alcohol (ethanol), 7 kinds of ketones (2-butanone, 2,3-pentanedione, 3-hydroxy-2-butanone, 1-hydroxy-2-propanone, furyl hydroxymethyl ketone, isophorone, 4-hydroxymethyl ketone) Methyl isophorone), 7 kinds of aldehydes (acetaldehyde, propionaldehyde, furfural, benzaldehyde, 5-methylfuranaldehyde, 2,3-dihydro-2,2,6-trimethylbenzaldehyde, n-undecanal), 4 kinds of esters (methyl formate, methyl acetate, vinyl acetate, furfuryl formate), 1 kind of alkanes (1-hexadecyl), 2 kinds of aromatic hydrocarbons (toluene, 1, 3,5-Trimethylbenzene), 1 type of acid (acetic acid), 3 types of furans (2-methylfuran, 2-ethylfuran, 2-n-pentylfuran), 1 type of thiophene (thiophene), pyrrole 4 types (N-methylpyrrole, pyrrole, 1-furfurylpyrrole, 2-acetylpyrrole), 1 type of pyridine (pyridine), 2 types of pyrazine (2,5-dimethylpyrazine, 2 , 6-dimethylpyrazine), 1 amide (formamide), 1 isocyanide (methyl hydrazine).

Table 10 Compounds related to the response changes of electronic nose during stir-frying of Gardenia

"-" means not detected.

The changes in the peak areas of compounds related to the electronic nose response changes during the frying process of Gardenia are shown in Table 11.

Table 11 Compounds related to the response changes of electronic nose during stir-frying of Gardenia

"a" represents a common peak; "-" represents no detection.

The relative percentage changes of various compounds during the frying process of Gardenia are shown in Figure 5. As shown in the figure, the odor change during the processing of gardenia is the comprehensive effect of multiple components, such as alcohols, ketones, aldehydes, esters, aromatic hydrocarbons, acids, furans, pyrroles, pyrazines, isotopes. The relative percentage of cyanogens in the whole frying process of gardenia is relatively high, and the average proportion is in the order of esters (12.08%)> aldehydes (10.19%)> ketones (9.91%)> Aromatic hydrocarbons (9.67%)>acids (9.28%)>alcohols (6.49%)>isocyanines (4.70%)>pyrazines (2.83%)>furans (2.79%)>pyrroles (1.54%) ), and showed the following trends: with the increase of frying time, the relative content of esters gradually increased; the relative content of ketones, aldehydes, aromatic hydrocarbons, acids, pyrroles and pyrazines increased first and then decreased; the relative contents of alcohols and isocyanides decreased gradually; the relative contents of furans decreased first and then increased. Therefore, it is speculated that these 10 kinds of substances may be the main component groups that cause the change of gardenia odor during the processing.

After further analysis of these types of compounds, it was found that during the frying process, the relative percentages exceeded 1% and the peak area changes showed an upward trend: methyl acetate, 2-methylfuran, and methyl alcohol; The decreasing trend is acetaldehyde, methyl formate, ethanol, toluene, 2-n-pentylfuran, 1-hydroxy-2-propanone, 2,5-dimethylpyrazine, 1,3,5-trimethylbenzene, acetic acid , furfural, pyrrole, isophorone and 4-methylene isophorone; propionaldehyde and 2-ethylfuran showed a trend of first decreasing and then increasing. These 18 compounds are the main compounds in the frying process of gardenia. characteristic ingredient. Among them, five components, methyl acetate, 2,5-dimethylpyrazine, acetic acid, furfural, and 4-methyleneisophorone, accounted for a relatively high percentage in the samples. Modern research has shown that isophorones have a mint or camphor-like smell. The aroma of gardenia overflows after frying, and its clear aroma turns into a burnt aroma in the later stage of frying. The change of its odor is related to the relative content of isophorone. changes are consistent. In the same way, the deepening of the coke aroma during the frying process of gardenia is related to the Medela reaction. For example, the relative content of furfural increases and then decreases with the increase of frying time, which is consistent with the change of gardenia odor.

The electronic nose is used to detect the odor data of 16 samples during the frying process, and according to the established odor digital standard and DFA model standard, it is determined that sample S1 is a raw product, S7 and S8 are fried products, and samples S11 and S12 They are coke products, samples S14 and S15 are charcoal products, and the rest are "too or too late". The specific results of the relative percentages of methyl acetate, 2,5-dimethylpyrazine, acetic acid, furfural, and 4-methyleneisophorone in each sample are shown in Table 11. For each component, the relative percentage content of raw gardenia S1 was taken as 1, and the ratio of the relative percentage content of the corresponding substances in the other samples was calculated. The specific results are shown in Table 11. According to the change ratio of substances shown in the table, the upper limit is increased by 10% and the lower limit is lowered by 10% to formulate the relative percentage change ratio limit of the five characteristic compounds.

Methyl acetate takes the relative percentage content of raw gardenia as 1, and calculates the ratio of the relative percentage content of fried gardenia, coke gardenia, gardenia charcoal and raw gardenia, fried gardenia, coke gardenia, and gardenia charcoal. Respectively control at 1.75-2.47, 1.74-2.71, 2.10-3.70; 2,5-dimethylpyrazine takes the relative percentage of raw gardenia as 1, and calculates fried gardenia, coke gardenia, gardenia charcoal and raw gardenia. The ratio of the relative percentage of gardenia, fried gardenia, coke gardenia, and gardenia charcoal should be controlled at 9.76-15.64, 4.19-8.15, 1.86-3.30 respectively; acetic acid is calculated based on the relative percentage of raw gardenia as 1. The ratio of the relative percentage of fried gardenia, coke gardenia, gardenia charcoal and raw gardenia, fried gardenia, coke gardenia, gardenia charcoal should be controlled at 5.13-7.40, 3.89-5.34, 2.48-3.48 respectively; furfural Taking the relative percentage content in raw gardenia as 1, calculate the ratio of the relative percentage content of fried gardenia, coke gardenia, gardenia charcoal and raw gardenia, fried gardenia, coke gardenia, and gardenia charcoal should be controlled at 246.60-397.65, 111.60-184.80, 61.20-101.75; 4-methylene isophorone takes the relative percentage of raw gardenia as 1, calculates the relative content of fried gardenia, coke gardenia, gardenia charcoal and raw gardenia Percentage ratio, fried gardenia, coke gardenia, gardenia charcoal should be controlled at 12.11-18.21, 14.11-18.81, 6.86-12.46 respectively. The specific results are shown in Table 12.

Table 12 Changes of characteristic compounds during frying

Claims (1)

1. A method for controlling the processing production degree and evaluating the quality of gardenia comprises the following steps: measuring odor value of fructus Gardeniae with electronic nose, analyzing volatile component content with gas chromatography-mass spectrometry, wherein the component content completely meets the specified range established by content limit method, and controlling the degree of fructus Gardeniae processing process and the quality of fructus Gardeniae processed product; the determination method comprises the following steps: the specified range is that the relative percentage content of methyl acetate, 2, 5-dimethyl pyrazine, acetic acid, furfural and 4-methylene isophorone is the evaluation index of processing gardenia: calculating the relative percentage ratio of parched fructus Gardeniae, fructus Gardeniae Preparata, and fructus Gardeniae charcoal to raw fructus Gardeniae by using methyl acetate as raw fructus Gardeniae relative percentage of 1, wherein the contents of parched fructus Gardeniae, fructus Gardeniae Preparata, and fructus Gardeniae charcoal should be respectively controlled at 1.75-2.47, 1.74-2.71, and 2.10-3.70; calculating the ratio of the relative percentage of fried gardenia, charred gardenia and gardenia charcoal to the relative percentage of raw gardenia by taking the relative percentage of raw gardenia as 1, wherein the relative percentage of fried gardenia, charred gardenia and gardenia charcoal is respectively controlled to be 9.76-15.64, 4.19-8.15 and 1.86-3.30; calculating the relative percentage ratio of fructus Gardeniae preparata, and fructus Gardeniae charcoal to fructus Gardeniae with acetic acid as raw fructus Gardeniae relative percentage of 1, wherein fructus Gardeniae preparata, and fructus Gardeniae charcoal should be controlled at 5.13-7.40, 3.89-5.34, and 2.48-3.48, respectively; calculating the ratio of the relative percentage of fried gardenia, charred gardenia and gardenia charcoal to the relative percentage of raw gardenia by taking the relative percentage of raw gardenia as 1, wherein the relative percentage of fried gardenia, charred gardenia and gardenia charcoal is respectively controlled to be 246.60-397.65, 111.60-184.80 and 61.20-101.75; the relative percentage of the 4-methylene isophorone in raw gardenia is 1, the ratio of the relative percentage of fried gardenia, charred gardenia and raw gardenia is calculated, and the relative percentage of the fried gardenia, the charred gardenia and the charred gardenia is respectively controlled to be 12.11-18.21, 14.11-18.81 and 6.86-12.46.

Images