CN113569470A - Fruit and vegetable respiration rate model parameter estimation method based on improved particle swarm optimization - Google Patents

Fruit and vegetable respiration rate model parameter estimation method based on improved particle swarm optimization Download PDF

Info

Publication number
CN113569470A
CN113569470A CN202110804035.0A CN202110804035A CN113569470A CN 113569470 A CN113569470 A CN 113569470A CN 202110804035 A CN202110804035 A CN 202110804035A CN 113569470 A CN113569470 A CN 113569470A
Authority
CN
China
Prior art keywords
particle
fruit
respiration rate
vegetable
rate model
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Granted
Application number
CN202110804035.0A
Other languages
Chinese (zh)
Other versions
CN113569470B (en
Inventor
曹乐
袁艳
李润
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Xian Technological University
Original Assignee
Xian Technological University
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Xian Technological University filed Critical Xian Technological University
Priority to CN202110804035.0A priority Critical patent/CN113569470B/en
Publication of CN113569470A publication Critical patent/CN113569470A/en
Application granted granted Critical
Publication of CN113569470B publication Critical patent/CN113569470B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F30/00—Computer-aided design [CAD]
    • G06F30/20—Design optimisation, verification or simulation
    • G06F30/27—Design optimisation, verification or simulation using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/004—Artificial life, i.e. computing arrangements simulating life
    • G06N3/006—Artificial life, i.e. computing arrangements simulating life based on simulated virtual individual or collective life forms, e.g. social simulations or particle swarm optimisation [PSO]

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Evolutionary Computation (AREA)
  • General Physics & Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • Software Systems (AREA)
  • Artificial Intelligence (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Biophysics (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Computer Hardware Design (AREA)
  • Health & Medical Sciences (AREA)
  • Medical Informatics (AREA)
  • Biomedical Technology (AREA)
  • Geometry (AREA)
  • Computational Linguistics (AREA)
  • Data Mining & Analysis (AREA)
  • General Health & Medical Sciences (AREA)
  • Molecular Biology (AREA)
  • Computing Systems (AREA)
  • Mathematical Physics (AREA)
  • Breeding Of Plants And Reproduction By Means Of Culturing (AREA)

Abstract

The invention discloses a fruit and vegetable respiration rate model parameter estimation method based on an improved particle swarm algorithm, which solves the problems that in the prior art, the accuracy of predicting the respiration rate of fruit and vegetable products is not high, the fitting effect is unstable only aiming at a single fruit and vegetable variety, provides a fruit and vegetable respiration rate model based on the influences of storage temperature, storage time and fruit and vegetable maturity, estimates fruit and vegetable respiration rate model parameters by improving the particle swarm algorithm, and has high accuracy and wide application range. The invention comprises the following steps: (1) constructing a fruit and vegetable respiration rate model; (2) initializing particle swarm parameters; (3) calculating a fitness value, and updating the particle speed and the particle position; (4) updating the global optimal particles; (5) updating the global worst particle; (6) adjusting the learning weight; (7) if the convergence condition or the maximum iteration number is reached, the next step is carried out, otherwise, the step (3) is returned; (8) and outputting the optimal value.

Description

Fruit and vegetable respiration rate model parameter estimation method based on improved particle swarm optimization
The technical field is as follows:
the invention belongs to the technical field of fruit and vegetable modified atmosphere packaging, and relates to a fruit and vegetable respiration rate model parameter estimation method based on an improved particle swarm algorithm.
Background art:
modified atmosphere packaging is the most effective fruit and vegetable fresh-keeping method at present. In the modified atmosphere packaging application of fruits and vegetables, two dynamic processes exist: on one hand, the picked fruits and vegetables still have breathActing to consume O in the package2Gradually reduce the concentration of the carbon dioxide to release CO2The concentration thereof was gradually increased. On the other hand, different gas partial pressures inside and outside the package lead to gas permeation. O outside the package2Higher than the concentration in the package, resulting in O2Permeation and diffusion into the package; and CO inside the package2Higher than the concentration outside the package, resulting in CO2The permeation diffuses to the outside of the package. When fruits and vegetables are consumed by respiration2At a rate equal to the penetration of O through the film2And fruit and vegetable breath to release CO2At a rate equal to the film bleeds CO2At a rate, build-up in the package with respect to O2And CO2And (5) regulating the atmosphere and dynamically balancing the environment.
The modified atmosphere packaging design is that the respiration rate of the fruits and vegetables is matched with the air permeability of the film to form the optimal storage condition, namely the gas dynamic balance environment. Under the condition of air-conditioned dynamic balance, the respiration rate of the fruits and vegetables is reduced to the lowest metabolic level, thereby prolonging the shelf life of the fruit and vegetable products.
The establishment of the fruit and vegetable respiration rate model is the basis for correctly designing modified atmosphere packaging. The poorly designed modified atmosphere packaging can not prolong the shelf life of fruit and vegetable products, but also can cause a series of problems such as anaerobic respiration, fermentation and the like, accelerate the spoilage of fruits and vegetables and even cause the problem of food quality safety. The factors influencing the respiration rate of the fruits and vegetables mainly comprise storage temperature, storage time and fruit and vegetable maturity. At present, no fruit and vegetable respiration model considering the comprehensive influence of storage temperature, storage time and fruit and vegetable maturity exists at home and abroad, and more Michaelies-Menten (Mie model) based on an enzyme kinetics theory and a respiration rate model based on statistics are used. However, both of these modeling methods have disadvantages: the Mie model cannot explain the respiration mechanism of the fruits and vegetables, and the accuracy of predicting the respiration rate of the fruit and vegetable products is not high; although the accuracy of the respiration rate model based on statistics is superior to that of a Mie model, the model generally adopts a multivariate function with gas concentration or temperature as an independent variable to calculate the respiration rate of fruits and vegetables, so that the model can only aim at a single fruit and vegetable variety, and the fitting effect is unstable.
The invention content is as follows:
the invention aims to provide a fruit and vegetable respiration rate model parameter estimation method based on an improved particle swarm algorithm, which solves the problems that in the prior art, the precision of predicting the respiration rate of fruit and vegetable products is not high, the fitting effect is unstable only aiming at a single fruit and vegetable variety, provides a fruit and vegetable respiration rate model based on the influences of storage temperature, storage time and fruit and vegetable maturity, further estimates fruit and vegetable respiration rate model parameters by improving the particle swarm algorithm, and has the advantages of high precision of the fruit and vegetable respiration rate model parameters and wide application range.
In order to achieve the purpose, the invention adopts the technical scheme that:
a fruit and vegetable respiration rate model parameter estimation method based on an improved particle swarm optimization is characterized in that: the method comprises the following steps:
(1) constructing a fruit and vegetable respiration rate model;
(2) initializing particle swarm parameters;
(3) calculating a fitness value, and updating the particle speed and the particle position;
(4) updating the global optimal particles;
(5) updating the global worst particle;
(6) adjusting the learning weight;
(7) if the convergence condition or the maximum iteration number is reached, the next step is carried out, otherwise, the step (3) is carried out;
(8) outputting an optimal value;
the method specifically comprises the following steps:
(1) a fruit and vegetable respiration rate model based on the joint influence of temperature, maturity and storage time is established as follows:
Figure BDA0003165699050000031
Figure BDA0003165699050000032
wherein the content of the first and second substances,
Figure BDA0003165699050000033
and
Figure BDA0003165699050000034
o under the conditions of storage temperature T, maturity Ma and storage time T2And CO2The respiration rate of (1), mL/(kg. h);
Figure BDA0003165699050000035
and
Figure BDA0003165699050000036
are each O2And CO2The basal respiration rate of (1), mL/(kg. h); alpha and beta are respectively O2And CO2Temperature coefficients of the breathing rate model; mu and nu are respectively O2And CO2A maturity coefficient of the respiration rate model; rho and sigma are respectively O2And CO2A storage time coefficient of the breathing rate model;
(2) the postharvest treatment is carried out on the fruits and vegetables, the hardness of the fruits and vegetables before and after the experiment is respectively measured by using a fruit hardness meter, the stem hardness of the green leaf vegetables is measured, and the hardness of the pulp tissues is measured for the rest of the green leaf vegetables. Measuring CO in modified atmosphere packaging under different storage temperature conditions2、O2Respectively calculating CO based on the gas concentrations2、O2The respiration rate of (a);
(3) construction of CO2、O2And solving the fruit and vegetable respiration rate model parameters by improving the particle swarm algorithm through the fitness function of the respiration rate model.
The post-harvest treatment and hardness measurement in the step (2) means that the picked fruits and vegetables are washed, air-dried and subjected to an experiment temperature TtestPrecooling for 30min, and measuring the hardness of the fruits and vegetables by using a fruit hardness meter;
measuring CO in the modified atmosphere package under the different storage temperature conditions in the step (2)2、O2Gas concentration and CO calculation2、O2The respiration rate of (A) is that a glass jar with a cover is selected, the glass jar is put in fruits and vegetables and then sealed, and a headspace gas analyzer is used through a small hole on the cover every 4 hoursMeasuring O in glass jar2And CO2Concentration, slight opening of the lid every other day to prevent O2Anaerobic respiration due to too low a concentration, time O2And CO2The respiration rate is calculated according to the formulas (3) and (4);
Figure BDA0003165699050000041
Figure BDA0003165699050000042
wherein the content of the first and second substances,
Figure BDA0003165699050000043
respectively about O at t moment2And CO2The respiration rate of (1), mL/(kg. h);
Figure BDA0003165699050000044
respectively at time t O2And CO2Concentration of (d)%;
Figure BDA0003165699050000045
respectively at an initial time O2And CO2Concentration of (d)%; vfHeadspace free volume, mL; t, tiRespectively as the current time and the initial time, h; w is the weight of the fruits and vegetables in kg.
The different storage temperature conditions in the step (2) refer to pre-cooling and experiment under the conditions of 5 ℃, 15 ℃ and 25 ℃.
The step (3) comprises the following steps:
s1, randomly initializing position vectors and velocity vectors of particles, setting the initial position of each particle as an individual optimal value, and setting the optimal value of a population as a global optimal value, wherein the particle swarm position vectors are parameters to be estimated by the fruit and vegetable respiration rate model
Figure BDA0003165699050000046
And
Figure BDA0003165699050000051
s2 construction of CO respectively2、O2Fitness function of the breathing rate model:
wherein, CO2The fitness function for the respiratory rate is as follows:
Figure BDA0003165699050000052
wherein, O2The fitness function for the respiratory rate is as follows:
Figure BDA0003165699050000053
s3, updating the speed and the position of each particle, calculating the fitness of the particles and evaluating a fitness function;
s4, constructing fitness matrixes FIT with the quantity of 0.1N according to the large-to-small orderingijWherein i is 1,2, …, 0.1N; j is 1,2, …, D, N is the number of particles, D is the particle dimension;
determining the selection probability of the global optimum vector element according to the following formula, selecting the particle FIT according to the probabilityijConstruction of gbest*Replacing the global optimal particles, and updating a global optimal value;
Figure BDA0003165699050000054
s5, sorting all the particle fitness from large to small, and randomly selecting two particles gbest from 0.1N particles with the fitness sorted from large to small for the particles with the worst fitness gwesta、gbestbAnd performing intersection according to the following formula to generate gworst to replace gworst;
Figure BDA0003165699050000055
wherein T is the total iteration number, T is the current iteration number, and epsilon is a random number of (0, 1);
s6, calculating the Euclidean distance between each particle and the global optimal particle, and calculating the Euclidean distance between each particle and the global optimal particle gbest according to the following formula
Figure BDA0003165699050000061
The learning weight is dynamically adjusted, updating each particle velocity according to the following formula:
Figure BDA0003165699050000062
s7, judging the termination condition of the algorithm, if the convergence of the algorithm meets the requirement or reaches the maximum iteration times, outputting a global optimum value
Figure BDA0003165699050000063
And
Figure BDA0003165699050000064
the algorithm ends. Otherwise, the Step is returned to Step 3.
In the step S3, the first step,
wherein the population X is composed of N particles { X ═ X1,x2,…xNAt the iteration time t, the position coordinate of the ith particle in the D-dimensional space is xi(t)＝(xi1,xi2,…,xiD) (ii) a Particle i velocity coordinate vi(t)＝(vi1,vi2,…,viD)；
Wherein the particle i experiences the best pbesti(t)＝(pbesyi1,pbesti2,…,pbestiD) As the individual optimum position; the best positions gbest (t) ═ pbest that all particles have experienced1(t),pbest2(t),…,pbestD(t)) as a global optimum position coordinate position;
wherein, position xi(t) and velocity vi(t) adjusting at time t +1 according to equations (7) and (8)And (3) finishing:
vij(t+1)＝wvij(t)+c1r1(pbestij(t)-xij(t))+c2r2(gbestj(t)-xij(t)) (7)
xij(t+1)＝xij(t)+vij(t+1) (8)
compared with the prior art, the invention has the advantages and effects that:
1. in order to determine the respiration rate of the fruits and vegetables, the invention provides a fruit and vegetable respiration rate model based on the influence of storage temperature, storage time and fruit and vegetable maturity; carrying out variation on globally optimal particles and globally worst particles, dynamically adjusting learning weight, and providing a wild goose particle swarm algorithm; and further estimating the fruit and vegetable respiration rate model parameters by improving the particle swarm optimization.
2. The fruit and vegetable respiration rate modeling method considers the influence of storage temperature, storage time and fruit and vegetable maturity on the respiration rate to carry out fruit and vegetable respiration rate modeling, and compared with the existing fruit and vegetable respiration rate model, the fruit and vegetable respiration rate modeling method is more in line with the change rule of the fruit and vegetable respiration rate. The improved particle swarm algorithm is simple in structure, easy to implement and high in convergence speed, and overcomes the defect that the particle swarm algorithm is easy to fall into a local optimal value. The fruit and vegetable respiration rate model fitted by improving the particle swarm algorithm has high parameter precision, wide application range and practical application value.
Description of the drawings:
FIG. 1 is a flow chart of a fruit and vegetable respiration rate model parameter estimation method according to the present invention;
FIG. 2 shows fruit and vegetable O in the above embodiment2A fitness function value convergence process of the respiratory rate;
FIG. 3 shows fruit and vegetable CO in the embodiment2And (5) convergence of fitness function values of the respiratory rate.
The specific implementation mode is as follows:
in order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention is described in further detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.
The invention relates to a fruit and vegetable respiration rate model parameter estimation method based on an improved particle swarm algorithm, which comprises the following steps of:
(1) starting;
(2) constructing a fruit and vegetable respiration rate model;
(3) initializing particle swarm parameters;
(4) calculating a fitness value, and updating the particle speed and the particle position;
(5) updating the global optimal particles;
(6) updating the global worst particle;
(7) adjusting the learning weight;
(8) if the convergence condition or the maximum iteration number is reached, the next step is carried out, otherwise, the step (4) is carried out;
(9) outputting an optimal value;
(10) and (6) ending.
The invention specifically comprises the following steps:
the method comprises the following steps of firstly, considering the influence of storage temperature, storage time and fruit and vegetable maturity on fruit and vegetable respiration rate change, and establishing a fruit and vegetable respiration rate model based on the joint influence of the temperature, the maturity and the storage time as follows:
Figure BDA0003165699050000081
Figure BDA0003165699050000082
wherein the content of the first and second substances,
Figure BDA0003165699050000083
and
Figure BDA0003165699050000084
o under the conditions of storage temperature T, maturity Ma and storage time T2And CO2The respiration rate of (1), mL/(kg. h);
Figure BDA0003165699050000085
and
Figure BDA0003165699050000086
are each O2And CO2The basal respiration rate of (1), mL/(kg. h); alpha and beta are respectively O2And CO2Temperature coefficients of the breathing rate model; mu and nu are respectively O2And CO2A maturity coefficient of the respiration rate model; rho and s are respectively O2And CO2Storage time coefficients of the breathing rate model.
And secondly, performing postharvest treatment on the fruits and vegetables, measuring the hardness of the fruits and vegetables before and after the experiment by using a fruit hardness meter, measuring the stem hardness of the green leaf vegetables, and measuring the hardness of the pulp tissues of the rest of the green leaf vegetables. Measuring CO in modified atmosphere packaging under different storage temperature conditions2、O2Respectively calculating CO based on the gas concentrations2、O2The respiratory rate of (c).
Third step, construction of CO2、O2Fitness function of the breathing rate model. Solving fruit and vegetable respiration rate model parameters by improving a particle swarm algorithm, comprising the following steps:
step1, randomly initializing position vectors and velocity vectors of the particles, setting the initial position of each particle as an individual optimal value, and setting the optimal value of the population as a global optimal value. And the particle swarm position vector is a parameter to be estimated by the fruit and vegetable respiration rate model. Wherein, the particle swarm position vector is the parameter to be estimated by the fruit and vegetable respiration rate model
Figure BDA0003165699050000091
And
Figure BDA0003165699050000092
step2, construction of CO separately2、O2Fitness function of the breathing rate model.
Among them, with respect to CO2The fitness function for the respiratory rate is as follows:
Figure BDA0003165699050000093
wherein, with respect to O2The fitness function for the respiratory rate is as follows:
Figure BDA0003165699050000094
and Step3, updating the speed and the position of each particle, calculating the fitness of the particles and evaluating the fitness function.
Step4, constructing fitness matrixes FIT with 0.1N number according to large-to-small orderingijWherein i is 1,2, …, 0.1N; j is 1,2, …, D, N is the number of particles, D is the particle dimension.
Determining the selection probability of the global optimum vector element according to the following formula, selecting the particle FIT according to the probabilityijConstruction of gbest*And replacing the global optimal particles and updating the global optimal value.
Figure BDA0003165699050000095
Step5, sorting all the particle fitness degrees from large to small, and randomly selecting two particles gbest from 0.1N particles with the fitness degrees sorted from large to small for the particles with the worst fitness degreea、gbestbAnd performing intersection according to the following formula to generate gworst to replace gworst.
Figure BDA0003165699050000101
Wherein T is the total iteration number, T is the current iteration number, and e is a random number of (0, 1). Step6, calculating the Euclidean distance between each particle and the global optimal particle, and calculating the Euclidean distance between each particle and the global optimal particle gbest according to the following formula
Figure BDA0003165699050000102
The learning weight is dynamically adjusted, updating each particle velocity according to the following formula:
Figure BDA0003165699050000103
step7, judging the termination condition of the algorithm, and outputting a global optimum value if the convergence of the algorithm meets the requirement or reaches the maximum iteration times
Figure BDA0003165699050000104
And
Figure BDA0003165699050000105
the algorithm ends. Otherwise, the Step is returned to Step 3.
Example (b):
referring to fig. 1, the invention provides a fruit and vegetable respiration rate model parameter estimation method based on an improved particle swarm optimization, the whole process mainly comprises a fruit and vegetable respiration rate model, the improved particle swarm optimization and model parameter estimation, and the method comprises the following steps:
firstly, establishing a fruit and vegetable respiration rate model based on the joint influence of temperature, maturity and storage time, wherein the formula (1) and the formula (2) are respectively O2And CO2The breathing rate of (c):
Figure BDA0003165699050000106
Figure BDA0003165699050000111
wherein the content of the first and second substances,
Figure BDA0003165699050000112
and
Figure BDA0003165699050000113
o under the conditions of storage temperature T, maturity Ma and storage time T2And CO2The respiration rate of (1), mL/(kg. h);
Figure BDA0003165699050000114
and
Figure BDA0003165699050000115
are each O2And CO2The basal respiration rate of (1), mL/(kg. h); alpha and beta are respectively O2And CO2Temperature coefficients of the breathing rate model; mu and nu are respectively O2And CO2A maturity coefficient of the respiration rate model; rho and s are respectively O2And CO2Storage time coefficients of the breathing rate model.
And secondly, performing postharvest treatment on the fruits and vegetables, measuring the hardness of the fruits and vegetables before and after the experiment by using a fruit hardness meter, measuring the stem hardness of the green leaf vegetables, and measuring the hardness of the pulp tissues of the rest of the green leaf vegetables. Measuring CO in modified atmosphere packaging under different storage temperature conditions2、O2Respectively calculating CO based on the gas concentrations2、O2The respiratory rate of (c).
Wherein, the post-harvest treatment and hardness measurement in the step two refers to that the picked fruits and vegetables are cleaned, air-dried and subjected to an experiment temperature TtestPrecooling for 30min, and measuring the hardness of the fruits and vegetables by using a fruit hardness meter. The measurement object in this example is a green leaf vegetable, and the stem hardness thereof is measured, and at the same time, the hardness of fresh freshly picked fruits and vegetables is measured by the above method, and is respectively recorded as Mf，kgf/cm2
Wherein the step two measures CO in the modified atmosphere package under the condition of different storage temperatures2、O2Gas concentration and CO calculation2、O2The respiration rate refers to that a glass tank with a cover is selected, fruits and vegetables are put in the glass tank and then sealed, and the oxygen content in the glass tank is measured by a headspace gas analyzer through a small hole on the cover every 4 hours2And CO2Concentration, slight opening of the lid every other day to prevent O2Anaerobic respiration due to too low a concentration, time O2And CO2The respiration rate is calculated by equations (3) and (4).
Figure BDA0003165699050000116
Figure BDA0003165699050000121
Wherein the content of the first and second substances,
Figure BDA0003165699050000122
respectively about O at t moment2And CO2The respiration rate of (1), mL/(kg. h);
Figure BDA0003165699050000123
respectively at time t O2And CO2Concentration of (d)%;
Figure BDA0003165699050000124
respectively at an initial time O2And CO2Concentration of (d)%; vfHeadspace free volume, mL; t, tiRespectively as the current time and the initial time, h; w is the weight of fruits and vegetables, kg
Wherein, the "different storage temperature conditions" in the second step means that pre-cooling and experiment are respectively carried out under the conditions of 5 ℃, 15 ℃ and 25 ℃.
Third step, construction of CO2、O2The fitness function of the respiration rate model solves the fruit and vegetable respiration rate model parameters by improving the particle swarm optimization, and comprises the following steps:
step1, randomly initializing position vectors and velocity vectors of the particles, setting the initial position of each particle as an individual optimal value, and setting the optimal value of the population as a global optimal value. Wherein, the position vector dimension of the particle swarm is 4, namely the parameters to be estimated by the fruit and vegetable respiration rate model
Figure BDA0003165699050000125
And
Figure BDA0003165699050000126
wherein, the population number N is 50, and the maximum iteration number T is 2000.
Step2, construction of CO separately2、O2Fitness function of the breathing rate model.
Among them, with respect to CO2The fitness function for the respiratory rate is as follows (5):
Figure BDA0003165699050000127
wherein, with respect to O2The fitness function of the respiration rate is as follows (6):
Figure BDA0003165699050000128
and Step3, updating the speed and the position of each particle, calculating the fitness of the particles and evaluating the fitness function.
Wherein the population X is composed of N particles { X ═ X1,x2,…xNAt the iteration time t, the position coordinate of the ith particle in the D-dimensional space is xi(t)＝(xi1,xi2,…,xiD) (ii) a Particle i velocity coordinate vi(t)＝(vi1,vi2,…,viD)；
Wherein the particle i experiences the best pbesti(t)＝(pbesyi1,pbesti2,…,pbestiD) As the individual optimum position. The best positions gbest (t) ═ pbest that all particles have experienced1(t),pbest2(t),…,pbestD(t)), as a global optimum position coordinate position.
Wherein, position xi(t) and velocity vi(t) at time t +1, the adjustment is performed according to equations (7) and (8):
vij(t+1)＝wvij(t)+c1r1(pbestij(t)-xij(t))+c2r2(gbestj(t)-xij(t)) (7)
xij(t+1)＝xij(t)+vij(t+1) (8)
step4, constructing fitness matrixes FIT with 0.1N number according to large-to-small orderingijWherein i is 1,2, xxx, 0.1N; j is 1,2, xxx, D, N is the population number, and D is the particle group position vector dimension.
Determining selection probability of global optimum vector element according to formula (9), selecting particle FIT according to probabilityijConstruction of gbest*And replacing the global optimal particles and updating the global optimal value.
Figure BDA0003165699050000131
Step5, sorting all the particle fitness degrees from large to small, and randomly selecting two particles gbest from 0.1N particles with the fitness degrees sorted from large to small for the particles with the worst fitness degreea、gbestbAnd performing intersection according to the following formula to generate gworst to replace gworst.
Figure BDA0003165699050000132
Wherein T is the maximum iteration number, T is the current iteration number, and epsilon is a random number of (0, 1).
Step6, calculating the Euclidean distance between each particle and the global optimal particle, and calculating the Euclidean distance between each particle and the global optimal particle gbest according to the formula (11)
Figure BDA0003165699050000141
The learning weight is dynamically adjusted and each particle velocity is updated according to equation (12):
Figure BDA0003165699050000142
step7, judgmentStopping the algorithm, if the fitness function converges to be less than epsilon or reaches the maximum iteration times, outputting a global optimum value
Figure BDA0003165699050000143
And
Figure BDA0003165699050000144
the algorithm ends. Otherwise, the Step is returned to Step 3.
FIG. 2 shows fruit and vegetable O in the above embodiment2The convergence process of the fitness function value of the respiration rate, and FIG. 3 shows the CO of the fruits and vegetables in the embodiment2And (5) convergence of fitness function values of the respiratory rate.
Experimental example:
in the embodiment, cucumber is taken as an experimental object, stored for 7 days at 5, 15 and 25 ℃, subjected to particle swarm parameter estimation and 2000-step iteration, and subjected to CO2Fitness function value of respiratory rate is reduced to 1.3, O2The fitness function value for the respiratory rate decreases to 2.9. Wherein, the experimental respiration rate and the predicted respiration rate are shown in table 1, and the estimated respiration rate model parameters are shown in table 2.
TABLE 1 respiration Rate values
Figure BDA0003165699050000151
Figure BDA0003165699050000161
TABLE 2 breathing Rate model parameters
Figure BDA0003165699050000162
The above description is only a preferred embodiment of the present invention, and is not intended to limit the scope of the present invention, and it should be noted that those skilled in the art should make modifications and variations without departing from the principle of the present invention.

Claims (6)

1. A fruit and vegetable respiration rate model parameter estimation method based on an improved particle swarm optimization is characterized in that: the method comprises the following steps:
(1) constructing a fruit and vegetable respiration rate model;
(2) initializing particle swarm parameters;
(3) calculating a fitness value, and updating the particle speed and the particle position;
(4) updating the global optimal particles;
(5) updating the global worst particle;
(6) adjusting the learning weight;
(7) if the convergence condition or the maximum iteration number is reached, the next step is carried out, otherwise, the step (3) is carried out;
(8) outputting an optimal value;
2. the fruit and vegetable respiration rate model parameter estimation method based on the improved particle swarm optimization algorithm according to claim 1, characterized in that:
the method comprises the following steps:
(1) a fruit and vegetable respiration rate model based on the joint influence of temperature, maturity and storage time is established as follows:
Figure FDA0003165699040000011
Figure FDA0003165699040000012
wherein the content of the first and second substances,
Figure FDA0003165699040000013
and
Figure FDA0003165699040000014
o under the conditions of storage temperature T, maturity Ma and storage time T2And CO2The respiration rate of (1), mL/(kg. h);
Figure FDA0003165699040000015
and
Figure FDA0003165699040000016
are each O2And CO2The basal respiration rate of (1), mL/(kg. h); alpha and beta are respectively O2And CO2Temperature coefficients of the breathing rate model; mu and nu are respectively O2And CO2A maturity coefficient of the respiration rate model; rho and sigma are respectively O2And CO2A storage time coefficient of the breathing rate model;
(2) the postharvest treatment is carried out on the fruits and vegetables, the hardness of the fruits and vegetables before and after the experiment is respectively measured by using a fruit hardness meter, the stem hardness of the green leaf vegetables is measured, and the hardness of the pulp tissues is measured for the rest of the green leaf vegetables. Measuring CO in modified atmosphere packaging under different storage temperature conditions2、O2Respectively calculating CO based on the gas concentrations2、O2The respiration rate of (a);
(3) construction of CO2、O2And solving the fruit and vegetable respiration rate model parameters by improving the particle swarm algorithm through the fitness function of the respiration rate model.
3. The fruit and vegetable respiration rate model parameter estimation method based on the improved particle swarm optimization algorithm according to claim 2, characterized in that:
the post-harvest treatment and hardness measurement in the step (2) means that the picked fruits and vegetables are washed, air-dried and subjected to an experiment temperature TtestPrecooling for 30min, and measuring the hardness of the fruits and vegetables by using a fruit hardness meter;
measuring CO in the modified atmosphere package under the different storage temperature conditions in the step (2)2、O2Gas concentration and CO calculation2、O2The respiration rate of (1) is that a glass tank with a cover is selected, the glass tank is put in fruits and vegetables and then sealed, and a headspace gas analyzer is used for measuring O in the glass tank through a small hole on the cover every 4 hours2And CO2Concentration of eachSlightly opening the lid every other day to prevent O2Anaerobic respiration due to too low a concentration, time O2And CO2The respiration rate is calculated according to the formulas (3) and (4);
Figure FDA0003165699040000021
Figure FDA0003165699040000022
wherein the content of the first and second substances,
Figure FDA0003165699040000031
respectively about O at t moment2And CO2The respiration rate of (1), mL/(kg. h);
Figure FDA0003165699040000032
respectively at time t O2And CO2Concentration of (d)%;
Figure FDA0003165699040000033
respectively at an initial time O2And CO2Concentration of (d)%; vfHeadspace free volume, mL; t, tiRespectively as the current time and the initial time, h; w is the weight of the fruits and vegetables in kg.
4. The fruit and vegetable respiration rate model parameter estimation method based on the improved particle swarm optimization algorithm according to claim 2, characterized in that: the different storage temperature conditions in the step (2) refer to pre-cooling and experiment under the conditions of 5 ℃, 15 ℃ and 25 ℃.
5. The fruit and vegetable respiration rate model parameter estimation method based on the improved particle swarm optimization algorithm according to claim 2, characterized in that:
the step (3) comprises the following steps:
s1 random initialization of particlesSetting the initial position of each particle as an individual optimal value and setting the optimal value of the population as a global optimal value, wherein the particle swarm position vector is a parameter to be estimated by the fruit and vegetable respiration rate model
Figure FDA0003165699040000034
And
Figure FDA0003165699040000035
s2 construction of CO respectively2、O2Fitness function of the breathing rate model:
wherein, CO2The fitness function for the respiratory rate is as follows:
Figure FDA0003165699040000036
wherein, O2The fitness function for the respiratory rate is as follows:
Figure FDA0003165699040000037
s3, updating the speed and the position of each particle, calculating the fitness of the particles and evaluating a fitness function;
s4, constructing fitness matrixes FIT with the quantity of 0.1N according to the large-to-small orderingijWherein i is 1,2, xxx, 0.1N; j is 1,2, …, D, N is the number of particles, D is the particle dimension;
determining the selection probability of the global optimum vector element according to the following formula, selecting the particle FIT according to the probabilityijConstruction of gbest*Replacing the global optimal particles, and updating a global optimal value;
Figure FDA0003165699040000041
s5, sorting all the particle fitness from large to smallFor the worst fitness particle gworst, two particle gbests are randomly selected from 0.1N particles with the fitness sorted from big to smalla、gbestbAnd performing intersection according to the following formula to generate gworst to replace gworst;
Figure FDA0003165699040000042
wherein T is the total iteration number, T is the current iteration number, and e is a random number of (0, 1);
s6, calculating the Euclidean distance between each particle and the global optimal particle, and calculating the Euclidean distance between each particle and the global optimal particle gbest according to the following formula
Figure FDA0003165699040000043
The learning weight is dynamically adjusted, updating each particle velocity according to the following formula:
Figure FDA0003165699040000044
s7, judging the termination condition of the algorithm, if the convergence of the algorithm meets the requirement or reaches the maximum iteration times, outputting a global optimum value
Figure FDA0003165699040000045
And
Figure FDA0003165699040000046
the algorithm ends. Otherwise, the Step is returned to Step 3.
6. The fruit and vegetable respiration rate model parameter estimation method based on the improved particle swarm optimization algorithm according to claim 5, characterized in that:
in the step S3, the first step,
wherein the population consists of N particlesSub-constituent population X ═ { X ═ X1,x2,…xNAt the iteration time t, the position coordinate of the ith particle in the D-dimensional space is xi(t)＝(xi1,xi2,…,xiD) (ii) a Particle i velocity coordinate vi(t)＝(vi1,vi2,…,viD)；
Wherein the particle i experiences the best pbesti(t)＝(pbesyi1,pbesti2,…,pbestiD) As the individual optimum position; the best positions gbest (t) ═ pbest that all particles have experienced1(t),pbest2(t),…,pbestD(t)) as a global optimum position coordinate position;
wherein, position xi(t) and velocity vi(t) at time t +1, the adjustment is performed according to equations (7) and (8):
vij(t+1)＝wvij(t)+c1r1(pbestij(t)-xij(t))+c2r2(gbestj(t)-xij(t)) (7)
xij(t+1)＝xij(t)+vij(t+1) (8)。
CN202110804035.0A 2021-07-16 2021-07-16 Fruit and vegetable respiration rate model parameter estimation method based on improved particle swarm optimization Active CN113569470B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202110804035.0A CN113569470B (en) 2021-07-16 2021-07-16 Fruit and vegetable respiration rate model parameter estimation method based on improved particle swarm optimization

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202110804035.0A CN113569470B (en) 2021-07-16 2021-07-16 Fruit and vegetable respiration rate model parameter estimation method based on improved particle swarm optimization

Publications (2)

Publication Number Publication Date
CN113569470A true CN113569470A (en) 2021-10-29
CN113569470B CN113569470B (en) 2024-04-05

Family

ID=78165048

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202110804035.0A Active CN113569470B (en) 2021-07-16 2021-07-16 Fruit and vegetable respiration rate model parameter estimation method based on improved particle swarm optimization

Country Status (1)

Country Link
CN (1) CN113569470B (en)

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102096732A (en) * 2011-01-07 2011-06-15 山东理工大学 Fruit and vegetable respiration rate modeling method
WO2017119987A1 (en) * 2016-01-07 2017-07-13 The Climate Corporation Generating digital models of crop yield based on crop planting dates and relative maturity values
CN107121407A (en) * 2017-06-02 2017-09-01 中国计量大学 The method that near-infrared spectrum analysis based on PSO RICAELM differentiates Cuiguan pear maturity
CN108399450A (en) * 2018-02-02 2018-08-14 武汉理工大学 Improvement particle cluster algorithm based on biological evolution principle
CN111462044A (en) * 2020-03-05 2020-07-28 浙江省农业科学院 Greenhouse strawberry detection and maturity evaluation method based on deep learning model

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102096732A (en) * 2011-01-07 2011-06-15 山东理工大学 Fruit and vegetable respiration rate modeling method
WO2017119987A1 (en) * 2016-01-07 2017-07-13 The Climate Corporation Generating digital models of crop yield based on crop planting dates and relative maturity values
CN107121407A (en) * 2017-06-02 2017-09-01 中国计量大学 The method that near-infrared spectrum analysis based on PSO RICAELM differentiates Cuiguan pear maturity
CN108399450A (en) * 2018-02-02 2018-08-14 武汉理工大学 Improvement particle cluster algorithm based on biological evolution principle
CN111462044A (en) * 2020-03-05 2020-07-28 浙江省农业科学院 Greenhouse strawberry detection and maturity evaluation method based on deep learning model

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
JIAN-SHUN CIOU等: "Determining Vital Signs with CW Doppler Radar Based on Particle Swarm Optimization", 《IEEE ASIA-PACIFIC MICROWAVE CONFERENCE》, pages 501 - 503 *
张立彬等: "基于ＩＰＳＯ－ＳＡ算法的温室番茄产量预测方法", 《浙江工业大学学报》, vol. 47, no. 5, pages 527 - 533 *

Also Published As

Publication number Publication date
CN113569470B (en) 2024-04-05

Similar Documents

Publication Publication Date Title
US10966435B2 (en) System for controlling low temperature injury of cold-sensitive fruit vegetables by combining intelligent pre-cooling and segmented controlled atmosphere storage
Yang et al. Modeling the effect of O2 and CO2 on respiration and quality of stored tomatoes
Jacxsens et al. Designing equilibrium modified atmosphere packages for fresh-cut vegetables subjected to changes in temperature
US10143210B2 (en) Dynamic control system and method for controlled atmosphere room
Torrieri et al. Modelling the respiration rate of minimally processed broccoli (Brassica rapa var. sylvestris) for modified atmosphere package design
CA3093864C (en) Ripening chamber and method for fruit ripening
CN109062296B (en) Grain storage method based on temperature and humidity monitoring
US20230048670A1 (en) Respiration ranking in controlled atmosphere rooms
Dadzie et al. Changes in respiration and ethylene production of apples in response to internal and external oxygen partial pressures
Nisbet et al. Accumulation and prediction of yeast assimilable nitrogen in New York winegrape cultivars
CN113569470A (en) Fruit and vegetable respiration rate model parameter estimation method based on improved particle swarm optimization
CN107079981B (en) Method for curing picked kiwi fruits
CN111280117B (en) Environment control method, device and system for keep-alive storage and transportation of aquatic products
CN111027784B (en) Method for predicting shelf life of cold fresh chicken
Taher et al. Respiration of barley seeds (Hordeum vulgare L.) under different storage conditions
Evelo et al. Modified atmosphere packaging of tomatoes; controlling gas and humidity
Alhamdan et al. Modeling of respiration rate of fresh date fruits (Barhi cultivar) under aerobic conditions
CN109393012A (en) A kind of air regulating fresh-keeping method saved for fresh fruit
Nisbet et al. Preharvest prediction of yeast assimilable nitrogen in Finger Lakes Riesling using linear and multivariate modeling
Lareo et al. Influence of temperature on shelf life of butterhead lettuce leaves under passive modified atmosphere packaging
Mwove et al. Microbial quality and shelf life prediction of vacuum-packaged ready to eat beef rounds containing gum arabic
KR101429715B1 (en) Apparatus for preserving agricultural products to maintain beneficial modified atmosphere in container by controlling opening/closing of vent hole through time-control, and method for preserving the same
CN110734866A (en) active freeze-dried powder for preventing and treating loquat postharvest diseases and preparation method and application thereof
Bessemans et al. A modelling approach to explain low apparent RQ-measurements of (D) CA stored Conference pear fruit
CN112070278B (en) Method for predicting roast duck shelf end point by combining electronic nose with principal component analysis

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant