CN115184395A - Fruit and vegetable weight loss rate prediction method and device, electronic equipment and storage medium - Google Patents
Fruit and vegetable weight loss rate prediction method and device, electronic equipment and storage medium Download PDFInfo
- Publication number
- CN115184395A CN115184395A CN202210579976.3A CN202210579976A CN115184395A CN 115184395 A CN115184395 A CN 115184395A CN 202210579976 A CN202210579976 A CN 202210579976A CN 115184395 A CN115184395 A CN 115184395A
- Authority
- CN
- China
- Prior art keywords
- weight loss
- loss rate
- fruits
- vegetables
- sample
- 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.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N25/00—Investigating or analyzing materials by the use of thermal means
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Health & Medical Sciences (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- General Physics & Mathematics (AREA)
- Computing Systems (AREA)
- Software Systems (AREA)
- Evolutionary Computation (AREA)
- Computational Linguistics (AREA)
- Molecular Biology (AREA)
- Biophysics (AREA)
- General Engineering & Computer Science (AREA)
- Biomedical Technology (AREA)
- Mathematical Physics (AREA)
- Data Mining & Analysis (AREA)
- Artificial Intelligence (AREA)
- Chemical & Material Sciences (AREA)
- Analytical Chemistry (AREA)
- Biochemistry (AREA)
- Immunology (AREA)
- Pathology (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention provides a method and a device for predicting the weight loss rate of fruits and vegetables, electronic equipment and a storage medium, wherein the method comprises the following steps: inputting the environmental temperature of the environment where the fruits and vegetables are located and the storage time of the fruits and vegetables into a trained neural network weight loss rate prediction model to obtain target weight loss rate prediction information of the fruits and vegetables; the trained neural network weight loss rate prediction model is obtained by training according to a first data group sample carrying a weight loss rate label; the first data set sample comprises a first storage temperature sample of the fruits and the vegetables and a first storage time sample corresponding to the first storage temperature sample. The method has accurate prediction result and simple process operation, can quickly and accurately dynamically monitor the weight loss rate information of the fruits and the vegetables in real time in the transportation process of the fruits and the vegetables, effectively improves the quality control capability of the fruits and the vegetables, and is favorable for realizing real-time tracking and management of the quality of the fruits and the vegetables in the storage and logistics transportation processes.
Description
Technical Field
The invention relates to the technical field of food safety detection, in particular to a method and a device for predicting the weight loss rate of fruits and vegetables, electronic equipment and a storage medium.
Background
The fruits and vegetables are used as fresh and perishable products, have extremely high nutritional values and are deeply favored by consumers. At present, the main marketing mode of fruits and vegetables is to transport the fruits and vegetables to a destination for sale in a short time after picking. However, in the transportation process of fruits and vegetables, the quality of the fruits and vegetables is very easily affected by the weather and the storage environment, and the weight loss rate of the fruits and vegetables is increased, the quality of the fruits and vegetables is reduced, and even the fruits and vegetables are rotten due to the respiration action, the water transpiration action and the like of the fruits and vegetables.
In order to realize the process detection of the quality changes of the fruits and vegetables such as the weight loss rate, the weight loss rate and the like are generally detected by adopting an off-line chemical test method in the prior art, however, the method is time-consuming and labor-consuming, and the real-time dynamic monitoring of the weight loss rate of the fruits and vegetables cannot be realized.
Therefore, how to efficiently and quickly monitor the weight loss rate of the fruits and vegetables in real time in the transportation process of the fruits and vegetables becomes a technical problem to be solved in the industry urgently.
Disclosure of Invention
The invention provides a method and a device for predicting the weight loss rate of fruits and vegetables, electronic equipment and a storage medium, which are used for dynamically monitoring the weight loss rate of the fruits and vegetables efficiently and quickly in the transportation process of the fruits and vegetables.
The invention provides a method for predicting the weight loss rate of fruits and vegetables, which comprises the following steps:
inputting the environmental temperature of the environment where the fruits and vegetables are located and the storage time of the fruits and vegetables into a trained neural network weight loss rate prediction model to obtain target weight loss rate prediction information of the fruits and vegetables;
the trained neural network weight loss rate prediction model is obtained by training according to a first data group sample carrying a weight loss rate label; the first data set sample comprises a first storage temperature sample of the fruits and the vegetables and a first storage time sample corresponding to the first storage temperature sample.
According to the fruit and vegetable weight loss rate prediction method provided by the invention, before the environmental temperature of the environment where the fruit and vegetable are located and the storage time of the fruit and vegetable are input into the trained neural network weight loss rate prediction model, the method further comprises the following steps:
obtaining a plurality of first data group samples and a weight loss rate label corresponding to each first data group sample; a plurality of first data set samples are determined based on the first storage temperature samples under a plurality of constant-temperature storage environments and a plurality of first storage time samples corresponding to each first storage temperature sample;
taking each first data group sample and the weight loss rate label corresponding to each first data group sample as a group of training samples to obtain a plurality of groups of training samples;
and training the neural network weight loss rate prediction model by using the multiple groups of training samples.
According to the fruit and vegetable weight loss rate prediction method provided by the invention, after the target weight loss rate prediction information of the fruit and vegetable is obtained, the method further comprises the following steps:
and determining the quality of the fruits and vegetables based on the target weight loss rate prediction information of the fruits and vegetables.
According to the fruit and vegetable weight loss rate prediction method provided by the invention, the training of the neural network weight loss rate prediction model by utilizing the multiple groups of training samples comprises the following steps:
for any group of training samples, inputting the training samples into the neural network weight loss rate prediction model to obtain weight loss rate prediction information corresponding to the training samples;
calculating a loss value according to the weight loss rate prediction information corresponding to the training sample and the weight loss rate label corresponding to the training sample by using a preset loss function;
and if the loss value is smaller than a preset threshold value, obtaining a trained neural network weight loss rate prediction model.
According to the fruit and vegetable weight loss rate prediction method provided by the invention, after the trained neural network weight loss rate prediction model is obtained, the method further comprises the following steps:
obtaining a plurality of second data group samples and a weight loss rate label corresponding to each second data group sample; the plurality of second data group samples are determined based on a plurality of second storage time samples in the same storage environment and a second storage temperature sample corresponding to each second storage time sample;
taking each second data group sample and the weight loss rate label corresponding to each second data group sample as a group of verification samples to obtain a plurality of groups of verification samples;
and performing model evaluation on the trained neural network weight loss rate prediction model by using the multiple groups of verification samples.
According to the fruit and vegetable weight loss rate prediction method provided by the invention, the neural network weight loss rate prediction model is formed on the basis of a radial basis function neural network, and network parameters of the radial basis function neural network at least comprise the maximum value of the number of neurons in a hidden layer and the number of neurons increased in two iterations;
the maximum value of the number of the neurons in the hidden layer ranges from 40 to 50, and the number of the neurons increased in the two iterations ranges from 1 to 3.
The invention also provides a device for predicting the fruit and vegetable weight loss rate, which comprises:
the prediction module is used for inputting the environmental temperature of the environment where the fruits and the vegetables are located and the storage time of the fruits and the vegetables into a trained neural network weight loss rate prediction model to obtain target weight loss rate prediction information of the fruits and the vegetables;
the trained neural network weight loss rate prediction model is obtained by training according to a first data group sample carrying a weight loss rate label; the first data set sample comprises a first storage temperature sample of the fruits and the vegetables and a first storage time sample corresponding to the first storage temperature sample.
The invention also provides electronic equipment, which comprises a memory, a processor and a computer program which is stored on the memory and can run on the processor, wherein the processor executes the program to realize the fruit and vegetable weight loss rate prediction method.
The invention also provides a non-transitory computer readable storage medium, on which a computer program is stored, wherein the computer program, when executed by a processor, implements any of the above-mentioned methods for predicting weight loss rate of fruits and vegetables.
The invention also provides a computer program product, which comprises a computer program, wherein when the computer program is executed by a processor, the method for predicting the weight loss rate of the fruits and vegetables is realized.
According to the fruit and vegetable weight loss rate prediction method, the fruit and vegetable weight loss rate prediction device, the trained neural network weight loss rate prediction model is obtained by training the first data group sample carrying the weight loss rate label, the first data group sample comprises the first storage temperature sample of the fruit and vegetable and the first storage time sample corresponding to the first storage temperature sample, the storage time and the environment temperature of the environment where the fruit and vegetable are located in the transportation process can be further obtained, the storage time and the environment temperature are input into the trained neural network weight loss rate prediction model, the target weight loss rate prediction information of the fruit and vegetable is obtained, the prediction result is accurate, the process operation is simple, the weight loss rate information of the fruit and vegetable can be rapidly and accurately monitored in real time in the transportation process of the fruit and vegetable, the fruit and vegetable quality control capacity is effectively improved, and the real-time tracking and management of the fruit and vegetable quality in the storage and logistics transportation process are facilitated.
Drawings
In order to more clearly illustrate the technical solutions of the present invention or the prior art, the drawings needed for the description of the embodiments or the prior art will be briefly described below, and it is obvious that the drawings in the following description are some embodiments of the present invention, and those skilled in the art can also obtain other drawings according to the drawings without creative efforts.
FIG. 1 is a schematic flow diagram of a fruit and vegetable weight loss rate prediction method provided by the invention;
FIG. 2 is a schematic structural diagram of a fruit and vegetable weight loss rate prediction device provided by the invention;
fig. 3 is a schematic structural diagram of an electronic device provided in the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention clearer, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings, and it is obvious that the described embodiments are some, but not all embodiments of the present invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
The method, the device, the electronic equipment and the storage medium for predicting the weight loss rate of the fruits and vegetables are described below with reference to fig. 1 to 3.
It should be noted that, during the logistics transportation process, the fruits and vegetables are easily affected by the weather temperature, especially the extreme fluctuation temperature, so that the respiration and water transpiration of the fruits and vegetables can be accelerated during the storage and transportation process, the weight loss rate of the fruits and vegetables can be rapidly increased, and the quality can be further reduced. However, in the prior art, the weight loss rate of the fruits and vegetables cannot be dynamically monitored in real time.
In order to solve the technical defects, the invention provides a fruit and vegetable weight loss rate prediction method, a device, electronic equipment and a storage medium.
Fig. 1 is a schematic flow diagram of a fruit and vegetable weight loss rate prediction method provided by the invention, as shown in fig. 1, including:
the trained neural network weight loss rate prediction model is obtained by training according to a first data group sample carrying a weight loss rate label; the first data set sample comprises a first storage temperature sample of the fruits and vegetables and a first storage time sample corresponding to the first storage temperature sample.
In particular, the fruits and vegetables described in the embodiments of the present invention refer to edible fruits and vegetables commercially available, which may specifically include grapefruit, grape, carrot, tomato, potato, and the like.
The storage time described in the embodiment of the invention refers to the time for storing the fruits and vegetables from the picking.
The environmental temperature of the embodiment of the invention refers to the temperature of the storage environment where the fruits and vegetables are stored, wherein the storage environment can be a refrigeration environment in the process of storage and transportation, and can also be the environment where the fruits and vegetables are located when the fruits and vegetables are sold in a retail site.
The target weight loss rate prediction information described in the embodiment of the invention refers to a predicted value of the weight loss rate of the fruits and vegetables correspondingly obtained based on the input environmental temperature of the environment where the fruits and vegetables are located and the storage time of the fruits and vegetables. Which can represent the weight loss rate level of fruits and vegetables. It can be understood that the output fruit and vegetable weight loss rate prediction information is different when different environmental temperatures and storage times are input. The fruit and vegetable weight loss rate prediction information can be specifically expressed in percentage, such as 5%.
The Neural Network weight loss rate prediction model described in the embodiment of the invention can be constructed based on a depth prediction Network, and is used for performing depth estimation and prediction on the weight loss rate of fruits and vegetables based on the environmental temperature of the environment where the fruits and vegetables are located and the storage time of the fruits and vegetables, wherein the depth prediction Network can be obtained based on the existing deep Neural Network training, and the input or output nonlinear system is identified by training the depth Neural Network model, so that the model has better stability, is suitable for measurement and calculation of the weight loss rate of the fruits and vegetables, and can adopt an RBFNN model if the advantages of the approximation capability, the classification capability, the fast learning convergence speed and the like of a Radial Basis Function Neural Network (RBFNN) are considered; it can also adopt neural networks such as Back Propagation (BP), long Short Term Memory (LSTM), and other neural networks which can be used for predicting the weight loss rate of fruits and vegetables.
The first storage temperature sample described in the embodiment of the invention refers to an environment temperature sample for constant-temperature storage of fruits and vegetables; the first storage time sample is a sample obtained from time data of fruit and vegetable storage at a temperature corresponding to the first storage temperature sample.
The first data set sample described in the embodiment of the invention comprises a first storage temperature sample of fruits and vegetables and a first storage time sample corresponding to the first storage temperature sample.
The trained neural network weight loss rate prediction model described in the embodiment of the invention is obtained by training according to a first data group sample carrying a weight loss rate label; the method is used for predicting the input environmental temperature and storage time data to be predicted, and learning the internal law of the influence of the environmental temperature and the storage time of the fruits and the vegetables on the weight loss rate of the fruits and the vegetables, thereby outputting accurate fruit and vegetable weight loss rate prediction information.
The weight loss rate label described in the present invention is predetermined according to the first data set sample to be predicted, and corresponds to the first data set sample one to one. That is, each first data group sample in the training samples is preset to carry a weight loss rate label corresponding to the first data group sample. In the embodiment of the present invention, the weight loss rate label can be obtained based on the experimental value of the weight loss rate of the fruits and vegetables under the first data group sample.
In the embodiment of the invention, the environmental temperature of the environment where the fruits and the vegetables are located and the storage time of the fruits and the vegetables can be monitored in real time based on the measuring equipment provided with the temperature sensor and the timer module, so that the environmental temperature of the environment where the fruits and the vegetables are located and the storage time of the fruits and the vegetables can be obtained in real time.
Further, in the embodiment of the invention, after the environmental temperature of the environment where the fruits and vegetables are located and the storage time of the fruits and vegetables are obtained, the environmental temperature of the environment where the fruits and vegetables are located and the storage time of the fruits and vegetables can be input into the trained neural network weight loss rate prediction model, and the trained neural network weight loss rate prediction model carries out deep estimation and prediction on the input environmental temperature and the storage time to obtain the target weight loss rate prediction information of the fruits and vegetables.
Based on the content of the above embodiment, as an optional embodiment, the neural network weight loss rate prediction model is constructed based on RBFNN, and network parameters of RBFNN at least include the maximum value of the number of neurons in the hidden layer and the number of neurons increased in two iterations;
the maximum value of the number of the neurons in the hidden layer ranges from 40 to 50, and the number of the neurons increased in the two iterations ranges from 1 to 3.
Specifically, in the embodiment of the invention, based on the advantages of the RBFNN model in the aspects of high approximation capability, classification capability, fast learning convergence speed and the like, the neural network weight loss rate prediction model is formed based on the RBFNN.
It is understood that the RBFNN model includes an input layer, a hidden layer, and an output layer.
Optionally, in an embodiment of the present invention, network parameters of the RBFNN may be set, where the set network parameters of the RBFNN at least include a maximum value of the number of neurons in the hidden layer and the number of neurons increased in two iterations, the maximum value of the number of neurons in the hidden layer may be set to 40, the number of neurons increased in two iterations in the model iteration process is set to 1, and in addition, a relative error of the RBFNN model may be set to 0.
In a specific embodiment, the fruits and vegetables are selected to be carrots, the first storage temperature samples of the carrots can be different temperatures, such as 36 ℃ (309K), 24 ℃ (297K), 12 ℃ (285K), 0 ℃ (273K), -10 ℃ (263K), -15 ℃ (258K), the first storage time samples and the corresponding weight loss rate labels of the first storage temperature samples and the first storage temperature samples are used as input layers of the RBFNN model, the weight loss rate prediction information of the carrots corresponding to the first storage temperature samples at different time points is used as output layers of the RBFNN model, and therefore the neural network weight loss rate prediction model is established and model training is carried out.
In the embodiment of the invention, by considering the advantages of the RBFNN model in the aspects of high approaching capacity, classification capacity, learning convergence speed and the like, the RBFNN model is adopted to construct the neural network weight loss rate prediction model, so that the prediction effect of the fruit and vegetable weight loss rate is improved, and the rapid and accurate real-time dynamic monitoring on the weight loss rate information of the fruits and vegetables is facilitated.
According to the fruit and vegetable weight loss rate prediction method provided by the embodiment of the invention, a trained neural network weight loss rate prediction model is obtained by training a first data set sample carrying a weight loss rate label, the first data set sample comprises a first storage temperature sample of fruits and vegetables and a first storage time sample corresponding to the first storage temperature sample, so that the storage time and the environmental temperature of the environment in the transportation process can be obtained, the storage time and the environmental temperature are input into the trained neural network weight loss rate prediction model, the target weight loss rate prediction information of the fruits and vegetables is obtained, the prediction result is accurate, the process operation is simple, the weight loss rate information of the fruits and vegetables can be rapidly and accurately monitored in real time in the transportation process of the fruits and vegetables, the quality control capability of the fruits and vegetables is effectively improved, and the real-time tracking and management of the fruit and vegetable quality in the storage and logistics transportation processes are facilitated.
Based on the content of the above embodiment, as an optional embodiment, after obtaining the prediction information of the target weight loss ratio of the fruit and vegetable, the method further includes:
and determining the quality of the fruits and vegetables based on the target weight loss rate prediction information of the fruits and vegetables.
It can be understood that the weight loss rate refers to the percentage of weight loss of the fruits and vegetables due to water evaporation caused by respiration and water transpiration, and can represent the quality of the fruits and vegetables. The weight loss ratio can be calculated according to the ratio of the weight loss of the fruits and vegetables before and after storage to the weight before storage. It can be understood that the larger the weight loss rate is, the poorer the quality of the fruits and vegetables is; conversely, the smaller the weight loss rate is, the better the quality of the fruits and vegetables is.
Specifically, in the embodiment of the invention, after the target weight loss rate prediction information of the fruits and vegetables is obtained, the quality of the fruits and vegetables can be judged based on the target weight loss rate prediction information of the fruits and vegetables. Generally, when the weight loss rate of the fruits and vegetables reaches 5%, the freshness of the fruits and vegetables is obviously reduced.
According to the method provided by the embodiment of the invention, the quality of the fruits and vegetables can be effectively determined by obtaining the target weight loss rate prediction information of the fruits and vegetables, and the quality loss of the fruits and vegetables can be rapidly monitored in real time under the extreme weather fluctuation temperature condition through logistics.
Based on the content of the above embodiment, as an optional embodiment, before inputting the environmental temperature of the environment where the fruits and vegetables are located and the storage time of the fruits and vegetables into the trained neural network weight loss ratio prediction model, the method further includes:
obtaining a plurality of first data group samples and a weight loss rate label corresponding to each first data group sample; the plurality of first data set samples are determined based on a plurality of first storage temperature samples under constant-temperature storage environment and a plurality of first storage time samples corresponding to each first storage temperature sample;
taking each first data group sample and the weight loss rate label corresponding to each first data group sample as a group of training samples to obtain a plurality of groups of training samples;
and (4) training the neural network weight loss rate prediction model by utilizing a plurality of groups of training samples.
Specifically, in the embodiment of the present invention, the plurality of first data set samples are determined and obtained based on the first storage temperature samples in the plurality of constant-temperature storage environments and the plurality of first storage time samples corresponding to each first storage temperature sample.
Illustratively, the purchased fruit and vegetable samples are stored in a plurality of constant temperature storage environments, such as 7 high precision constant temperature refrigerators with storage temperatures of 36 + -0.5 deg.C, 24 + -0.5 deg.C, 12 + -0.5 deg.C, 0 + -0.5 deg.C, -5 + -0.5 deg.C, -10 + -0.5 deg.C, and-15 + -0.5 deg.C, respectively. And (3) acquiring the weight loss rate information of the fruit and vegetable sample once every 12 hours as a corresponding weight loss rate label by a chemical test method. At this time, it can be understood that the first storage temperature sample is 36 ± 0.5 ℃, 24 ± 0.5 ℃, 12 ± 0.5 ℃, 0 ± 0.5 ℃, -5 ± 0.5 ℃, -10 ± 0.5 ℃ or-15 ± 0.5 ℃, and when the first storage temperature sample is 36 ± 0.5 ℃, the first storage time sample corresponding to the first storage temperature sample is a time sample of 0 hour, 12 hours, 24 hours, 36 hours or 48 hours and the like at 36 ± 0.5 ℃; when the first storage temperature sample is 0 +/-0.5 ℃, the first storage time sample corresponding to the first storage temperature sample is a time sample of 0 hour, 12 hours, 24 hours, 36 hours or 48 hours and the like at the temperature of 0 +/-0.5 ℃; when the first storage temperature sample is at minus 10 +/-0.5 ℃, the first storage time sample corresponding to the first storage temperature sample is a time sample of 0 hour, 12 hours, 24 hours, 36 hours or 48 hours and the like at minus 10 +/-0.5 ℃; similarly, the first storage time samples corresponding to the other 4 first storage temperature samples can be obtained, so that the weight loss rate labels corresponding to all 7 first data group samples and each first data group sample can be obtained.
Further, in the embodiment of the present invention, each first data group sample and the weight loss rate label corresponding to each first data group sample are used as a set of training samples, that is, each first data group sample with a weight loss rate label is used as a set of training samples, so that multiple sets of training samples can be obtained.
In an embodiment of the present invention, the first data set samples are in one-to-one correspondence with the weight loss rate labels carried by the first data set samples.
Further, after a plurality of groups of training samples are obtained, the plurality of groups of training samples are sequentially input into the neural network weight loss rate prediction model, namely, the first data group sample and the carried weight loss rate label in each group of training samples are simultaneously input into the neural network weight loss rate prediction model, model parameters in the neural network weight loss rate prediction model are adjusted according to each output result of the neural network weight loss rate prediction model by calculating a loss function value, and finally the training process of the neural network weight loss rate prediction model is completed.
According to the method provided by the embodiment of the invention, each first data group sample and the weight loss rate label corresponding to each first data group sample are used as a group of training samples to obtain a plurality of groups of training samples, so that the neural network weight loss rate prediction model can be effectively trained by using the plurality of groups of training samples, and the precision of the trained model is improved.
Based on the content of the foregoing embodiment, as an optional embodiment, the training of the neural network weight loss ratio prediction model by using multiple groups of training samples includes:
for any group of training samples, inputting the training samples into a neural network weight loss rate prediction model to obtain weight loss rate prediction information corresponding to the training samples;
calculating a loss value according to the weight loss rate prediction information corresponding to the training sample and the weight loss rate label corresponding to the training sample by using a preset loss function;
and if the loss value is smaller than the preset threshold value, obtaining a trained neural network weight loss rate prediction model.
Specifically, the preset loss function described in the embodiment of the present invention refers to a loss function preset in the neural network weight loss ratio prediction model, and is used for model evaluation.
The preset threshold described in the embodiment of the invention refers to a threshold preset by the model, and is used for obtaining the minimum loss value and completing the model training.
After a plurality of groups of training samples are obtained, for any group of training samples, a first data group sample in the training samples and a weight loss rate label carried by the first data group sample are simultaneously input into a neural network weight loss rate prediction model, and weight loss rate prediction information corresponding to the training samples is output.
On the basis, a preset loss function is utilized to calculate a loss value according to the weight loss rate prediction information corresponding to the training sample and the real weight loss rate label in the training sample. For example, the weight loss rate label can be expressed as a one-hot vector, and the preset loss function can adopt a cross entropy loss function.
In the embodiment of the present invention, the expression mode of the weight loss rate label and the preset loss function may be set according to actual requirements, which is not specifically limited herein.
After the loss value is obtained through calculation, the training process is finished, network parameters in the neural network weight loss rate prediction model are updated through an error back propagation algorithm, and then the next training is carried out. In the training process, if the loss value obtained by calculation aiming at a certain training sample is smaller than a preset threshold value or reaches a preset maximum iteration number, the training of the neural network weight loss rate prediction model is finished.
According to the method provided by the embodiment of the invention, the neural network weight loss rate prediction model is trained, and the loss value of the neural network weight loss rate prediction model is controlled within a preset range, so that the accuracy of the neural network weight loss rate prediction model for predicting the weight loss rate is improved.
Based on the content of the foregoing embodiment, as an optional embodiment, after obtaining the trained neural network weight loss ratio prediction model, the method further includes:
obtaining a plurality of second data group samples and a weight loss rate label corresponding to each second data group sample; the plurality of second data set samples are determined based on a plurality of second storage time samples and a second storage temperature sample corresponding to each second storage time sample in the same storage environment;
taking each second data group sample and the weight loss rate label corresponding to each second data group sample as a group of verification samples to obtain a plurality of groups of verification samples;
and performing model evaluation on the trained neural network weight loss rate prediction model by using a plurality of groups of verification samples.
In particular, the second storage time sample described in the embodiments of the present invention refers to a time sample obtained based on continuous time data in the same storage environment.
The second storage temperature samples described in the embodiments of the present invention may be storage temperature samples corresponding to each second storage time sample.
It is understood that the second data set samples include a second storage time sample and a second storage temperature sample corresponding to the second storage time sample.
In the embodiment of the invention, the plurality of second data group samples are determined based on the plurality of second storage time samples in the same storage environment and the second storage temperature sample corresponding to each second storage time sample, that is, in the same storage environment, the storage temperature of the fruits and vegetables may change along with the time duration, thereby forming the fluctuating temperature storage condition. As shown in table 1, the second storage time may be any one of 0h to 48h, and when the second storage time is 0h, the corresponding second storage temperature is 273K; when the second storage time is between 0 and 12h, the corresponding second storage temperature is still 273K; however, the corresponding second storage temperature is 263K for a second storage time of between 12h and 24h, 273K for a second storage time of between 24h and 36h, and 285K for a second storage time of between 36h and 48h, wherein h is h and K is K.
TABLE 1
Further, in the embodiment of the present invention, each second data group sample and the weight loss rate label corresponding to each second data group sample are used as a set of training samples, that is, each first data group sample with a weight loss rate label is used as a set of verification samples, so that multiple sets of verification samples can be obtained.
It can be understood that, in this embodiment, the second data group samples have a one-to-one correspondence with the weight loss rate labels carried by them.
Further, after obtaining a plurality of groups of verification samples, performing model evaluation on the trained neural network weight loss rate prediction model by using the plurality of groups of verification samples, namely sequentially inputting the plurality of groups of verification samples into the trained neural network weight loss rate prediction model, simultaneously inputting the first data group samples and the carried weight loss rate labels in each group of training samples into the trained neural network weight loss rate prediction model, obtaining each output result of the trained neural network weight loss rate prediction model, and calculating a correlation coefficient R between the weight loss rate labels (namely experimental values or true values) of the fruits and vegetables and corresponding fruit and vegetable weight loss rate predicted values 2 Wherein the correlation coefficient R 2 Is expressed as follows:
wherein the content of the first and second substances,
indicates the predicted value, y i The actual value is represented by the value of,
mean values representing the true values;
therefore, model evaluation of the trained neural network weight loss rate prediction model is completed.
In the embodiment of the invention, the fitting confidence interval calculation of the predicted value and the actual value of the fruit and vegetable weight loss rate is carried out under the storage condition of fluctuating temperature, the weight loss rate prediction information is obtained based on the trained neural network weight loss rate prediction model, the predicted value of the fruit and vegetable weight loss rate is determined, and the correlation coefficient R between the actual value and the predicted value of the fruit and vegetable weight loss rate is calculated 2 >0.97, and each prediction result falls in a 90% prediction band, which can show that the confidence coefficient of the neural network weight loss rate prediction model trained by the invention is more than 90%. Meanwhile, under the storage condition of fluctuating temperature, the relative error between the actual value of the fruit and vegetable weight loss rate and the predicted value of the fruit and vegetable weight loss rate can be calculated to be within +/-16 percent, and the prediction model can be accepted by considering the inherent difference of fruit and vegetable samples and the complexity of dynamic reaction.
According to the method provided by the embodiment of the invention, the neural network weight loss rate prediction model is evaluated, and the accuracy of the neural network weight loss rate prediction model is evaluated by adopting the correlation coefficient, so that the prediction accuracy of the neural network weight loss rate prediction model is further improved.
The fruit and vegetable weight loss rate prediction method provided by the embodiment of the invention is based on fruit and vegetable weight loss rate measurement and is based on the practical application aspect of not generating damage to fruits and vegetables, a method for rapidly and real-timely monitoring the quality loss of fruits and vegetables under the condition of extreme weather fluctuation temperature by supporting logistics is constructed, the fruit and vegetable detection method can be expanded, quality tracing and management of the fruits and vegetables in the storage and logistics transportation processes can be conveniently carried out by quality inspectors and the like, and the fruit and vegetable quality control technology is improved. Based on the trained neural network weight loss rate prediction model provided by the invention, related fruit and vegetable weight loss rate real-time monitoring equipment can be customized. Specific effects can be explained from the following embodiments.
In the first embodiment, a quality inspector detects the weight loss rate change detection process of fruits and vegetables from the beginning of distribution to a sales site in real time:
firstly, placing fruits and vegetables in a logistics carriage, starting fruit and vegetable weight loss rate real-time monitoring equipment when logistics distribution starts, and checking real-time weight loss rate data change;
then, after system equipment is started, data acquisition and processing are carried out on the storage time of the fruits and the vegetables and the environmental temperature of the environment in which the fruits and the vegetables are located in the transportation process, and real-time detection data of the weight loss rate of the fruits and the vegetables are displayed;
finally, the weight loss rate change of the fruits and vegetables is monitored in real time through the equipment, and accordingly, the quality inspector determines the quality loss change of the batch of fruits and vegetables at the time point.
In the second embodiment, the weight loss rate change detection process of the consumer in the fruit and vegetable retail stage is detected in real time:
firstly, after the fruits and vegetables are sold in places such as supermarket markets and the like, each retail site is provided with customized fruit and vegetable weight loss rate real-time monitoring equipment and a customized fruit and vegetable weight loss rate real-time monitoring system for merchants and consumers;
then, the consumer determines the quality of the fruit and vegetable by observing the current fruit and vegetable weight loss rate displayed by the equipped fruit and vegetable weight loss rate real-time monitoring equipment;
finally, the quality of the fruits and vegetables is detected through the equipment, and whether the fruits and vegetables are purchased or not is determined by a consumer.
The method provided by the embodiment of the invention can accurately simulate the fruit and vegetable weight loss rate change of logistics transportation in the extreme weather fluctuating temperature process, can effectively improve the fruit and vegetable weight loss rate monitoring capability of the logistics transportation in the fluctuating temperature process, ensures the fruit and vegetable quality, helps quality inspectors to quickly detect the fruit and vegetable quality in real time, can accurately predict the weight loss rate change information of the fruit and vegetable in the shelf retail stage, and provides quality basis for consumers.
The fruit and vegetable weight loss rate prediction device provided by the invention is described below, and the fruit and vegetable weight loss rate prediction device described below and the fruit and vegetable weight loss rate prediction method described above can be correspondingly referred to.
Fig. 2 is a schematic structural diagram of a device for predicting a weight loss rate of fruits and vegetables, as shown in fig. 2, including:
the prediction module 210 is used for inputting the environmental temperature of the environment where the fruits and vegetables are located and the storage time of the fruits and vegetables into the trained neural network weight loss rate prediction model to obtain target weight loss rate prediction information of the fruits and vegetables;
the trained neural network weight loss rate prediction model is obtained by training according to a first data group sample carrying a weight loss rate label; the first data set sample comprises a first storage temperature sample of the fruits and the vegetables and a first storage time sample corresponding to the first storage temperature sample.
The fruit and vegetable weight loss rate prediction device described in this embodiment may be used to implement the above fruit and vegetable weight loss rate prediction method embodiment, and the principle and technical effect are similar, which are not described herein again.
The fruit and vegetable weight loss rate prediction device provided by the embodiment of the invention obtains the trained neural network weight loss rate prediction model by training the first data set sample carrying the weight loss rate label, wherein the first data set sample comprises the first storage temperature sample of the fruit and vegetable and the first storage time sample corresponding to the first storage temperature sample, so that the storage time and the environmental temperature of the environment in the transportation process can be obtained, and the storage time and the environmental temperature are input into the trained neural network weight loss rate prediction model to obtain the target weight loss rate prediction information of the fruit and vegetable.
Fig. 3 is a schematic structural diagram of an electronic device provided in the present invention, and as shown in fig. 3, the electronic device may include: a processor (processor) 310, a communication Interface (communication Interface) 320, a memory (memory) 330 and a communication bus 340, wherein the processor 310, the communication Interface 320 and the memory 330 communicate with each other via the communication bus 340. The processor 310 may call the logic instructions in the memory 330 to execute the method for predicting the weight loss rate of fruits and vegetables provided by the above methods, where the method includes: inputting the environmental temperature of the environment where the fruits and vegetables are located and the storage time of the fruits and vegetables into a trained neural network weight loss rate prediction model to obtain target weight loss rate prediction information of the fruits and vegetables; the trained neural network weight loss rate prediction model is obtained by training according to a first data group sample carrying a weight loss rate label; the first data set sample comprises a first storage temperature sample of the fruits and the vegetables and a first storage time sample corresponding to the first storage temperature sample.
In addition, the logic instructions in the memory 330 may be implemented in the form of software functional units and stored in a computer readable storage medium when the software functional units are sold or used as independent products. Based on such understanding, the technical solution of the present invention may be embodied in the form of a software product, which is stored in a storage medium and includes instructions for causing a computer device (which may be a personal computer, a server, or a network device) to execute all or part of the steps of the method according to the embodiments of the present invention. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a magnetic disk, or an optical disk, and various media capable of storing program codes.
In another aspect, the present invention further provides a computer program product, where the computer program product includes a computer program, the computer program may be stored on a non-transitory computer readable storage medium, and when the computer program is executed by a processor, a computer can execute the fruit and vegetable weight loss rate prediction method provided by the above methods, where the method includes: inputting the environmental temperature of the environment where the fruits and vegetables are located and the storage time of the fruits and vegetables into a trained neural network weight loss rate prediction model to obtain target weight loss rate prediction information of the fruits and vegetables; the trained neural network weight loss rate prediction model is obtained by training according to a first data group sample carrying a weight loss rate label; the first data set sample comprises a first storage temperature sample of the fruits and vegetables and a first storage time sample corresponding to the first storage temperature sample.
In another aspect, the present invention further provides a non-transitory computer-readable storage medium, on which a computer program is stored, where the computer program, when executed by a processor, is implemented to perform the method for predicting weight loss rate of fruit and vegetable provided by the above methods, where the method includes: inputting the environmental temperature of the environment where the fruits and vegetables are located and the storage time of the fruits and vegetables into a trained neural network weight loss rate prediction model to obtain target weight loss rate prediction information of the fruits and vegetables; the trained neural network weight loss rate prediction model is obtained by training according to a first data group sample carrying a weight loss rate label; the first data set sample comprises a first storage temperature sample of the fruits and vegetables and a first storage time sample corresponding to the first storage temperature sample.
The above-described embodiments of the apparatus are merely illustrative, and the units described as separate parts may or may not be physically separate, and parts displayed as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of the present embodiment. One of ordinary skill in the art can understand and implement it without inventive effort.
Through the above description of the embodiments, those skilled in the art will clearly understand that each embodiment can be implemented by software plus a necessary general hardware platform, and certainly can also be implemented by hardware. With this understanding in mind, the above-described technical solutions may be embodied in the form of a software product, which can be stored in a computer-readable storage medium such as ROM/RAM, magnetic disk, optical disk, etc., and includes instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute the methods described in the embodiments or some parts of the embodiments.
Finally, it should be noted that: the above examples are only intended to illustrate the technical solution of the present invention, but not to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; and such modifications or substitutions do not depart from the spirit and scope of the corresponding technical solutions of the embodiments of the present invention.
Claims (10)
1. A fruit and vegetable weight loss rate prediction method is characterized by comprising the following steps:
inputting the environmental temperature of the environment where the fruits and vegetables are located and the storage time of the fruits and vegetables into a trained neural network weight loss rate prediction model to obtain target weight loss rate prediction information of the fruits and vegetables;
the trained neural network weight loss rate prediction model is obtained by training according to a first data group sample carrying a weight loss rate label; the first data set sample comprises a first storage temperature sample of the fruits and the vegetables and a first storage time sample corresponding to the first storage temperature sample.
2. The method for predicting the fruit and vegetable weight loss rate according to claim 1, wherein before the environmental temperature of the environment where the fruit and vegetable are located and the storage time of the fruit and vegetable are input into the trained neural network weight loss rate prediction model, the method further comprises the following steps:
obtaining a plurality of first data group samples and a weight loss rate label corresponding to each first data group sample; a plurality of first data set samples are determined based on the first storage temperature samples under a plurality of constant-temperature storage environments and a plurality of first storage time samples corresponding to each first storage temperature sample;
taking each first data group sample and the weight loss rate label corresponding to each first data group sample as a group of training samples to obtain a plurality of groups of training samples;
and training the neural network weight loss rate prediction model by using the plurality of groups of training samples.
3. The method for predicting the fruit and vegetable weight loss rate according to claim 1, wherein after the information for predicting the target weight loss rate of the fruit and vegetable is obtained, the method further comprises the following steps:
and determining the quality of the fruits and vegetables based on the target weight loss rate prediction information of the fruits and vegetables.
4. The fruit and vegetable weight loss rate prediction method according to claim 2, wherein the training of the neural network weight loss rate prediction model by using the plurality of groups of training samples comprises:
for any group of training samples, inputting the training samples into the neural network weight loss rate prediction model to obtain weight loss rate prediction information corresponding to the training samples;
calculating a loss value according to the weight loss rate prediction information corresponding to the training sample and the weight loss rate label corresponding to the training sample by using a preset loss function;
and if the loss value is smaller than a preset threshold value, obtaining a trained neural network weight loss rate prediction model.
5. The fruit and vegetable weight loss rate prediction method according to claim 4, characterized in that after the trained neural network weight loss rate prediction model is obtained, the method further comprises the following steps:
obtaining a plurality of second data group samples and a weight loss rate label corresponding to each second data group sample; the plurality of second data group samples are determined based on a plurality of second storage time samples in the same storage environment and a second storage temperature sample corresponding to each second storage time sample;
taking each second data group sample and the weight loss rate label corresponding to each second data group sample as a group of verification samples to obtain a plurality of groups of verification samples;
and performing model evaluation on the trained neural network weight loss rate prediction model by using the multiple groups of verification samples.
6. The fruit and vegetable weight loss rate prediction method according to any one of claims 1-5, characterized in that the neural network weight loss rate prediction model is formed based on a radial basis function neural network, and network parameters of the radial basis function neural network at least comprise the maximum value of the number of neurons in the hidden layer and the number of neurons increased in two iterations;
the maximum value of the number of the neurons in the hidden layer ranges from 40 to 50, and the number of the neurons increased in the two iterations ranges from 1 to 3.
7. A fruit and vegetable weight loss rate prediction device is characterized by comprising:
the prediction module is used for inputting the environmental temperature of the environment where the fruits and vegetables are located and the storage time of the fruits and vegetables into the trained neural network weight loss rate prediction model to obtain target weight loss rate prediction information of the fruits and vegetables;
the trained neural network weight loss rate prediction model is obtained by training according to a first data group sample carrying a weight loss rate label; the first data set sample comprises a first storage temperature sample of the fruits and the vegetables and a first storage time sample corresponding to the first storage temperature sample.
8. An electronic device comprising a memory, a processor and a computer program stored on the memory and operable on the processor, wherein the processor implements the method for predicting weight loss rate of fruits and vegetables according to any one of claims 1 to 6 when executing the program.
9. A non-transitory computer readable storage medium, having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the method for predicting weight loss rate of fruit and vegetable according to any one of claims 1 to 6.
10. A computer program product comprising a computer program, wherein the computer program, when executed by a processor, implements the method for predicting weight loss rate of fruits and vegetables according to any one of claims 1 to 6.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210579976.3A CN115184395A (en) | 2022-05-25 | 2022-05-25 | Fruit and vegetable weight loss rate prediction method and device, electronic equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210579976.3A CN115184395A (en) | 2022-05-25 | 2022-05-25 | Fruit and vegetable weight loss rate prediction method and device, electronic equipment and storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115184395A true CN115184395A (en) | 2022-10-14 |
Family
ID=83512942
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210579976.3A Pending CN115184395A (en) | 2022-05-25 | 2022-05-25 | Fruit and vegetable weight loss rate prediction method and device, electronic equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115184395A (en) |
Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107966512A (en) * | 2017-11-24 | 2018-04-27 | 浙江省农业科学院 | Biogenic amine Forecasting Methodology in Larimichthys crocea based on BP artificial neural networks |
| CN108304970A (en) * | 2018-02-05 | 2018-07-20 | 西北农林科技大学 | The method for quick predicting and system of Apple, air conditioned storage monitoring system |
| CN108898218A (en) * | 2018-05-24 | 2018-11-27 | 阿里巴巴集团控股有限公司 | A kind of training method of neural network model, device and computer equipment |
| CN109816150A (en) * | 2018-12-29 | 2019-05-28 | 中国农业大学 | One kind adopting rear Table Grape forecasting shelf life method and device |
| CN109840854A (en) * | 2019-03-04 | 2019-06-04 | 中国科学技术大学 | Grain mildew prediction technique and device |
| CN110210680A (en) * | 2019-06-11 | 2019-09-06 | 北京农业信息技术研究中心 | A kind of fish body Noninvasive Measuring Method of Freshness and device based on temperature change |
| WO2020140377A1 (en) * | 2019-01-04 | 2020-07-09 | 平安科技（深圳）有限公司 | Neural network model training method and apparatus, computer device, and storage medium |
| CN113433081A (en) * | 2021-07-06 | 2021-09-24 | 江南大学 | Dynamic neural network-based quality detection method and system in fruit and vegetable drying process |
| CN113791055A (en) * | 2021-08-17 | 2021-12-14 | 北京农业信息技术研究中心 | Fish freshness detection method and system |
-
2022
- 2022-05-25 CN CN202210579976.3A patent/CN115184395A/en active Pending
Patent Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107966512A (en) * | 2017-11-24 | 2018-04-27 | 浙江省农业科学院 | Biogenic amine Forecasting Methodology in Larimichthys crocea based on BP artificial neural networks |
| CN108304970A (en) * | 2018-02-05 | 2018-07-20 | 西北农林科技大学 | The method for quick predicting and system of Apple, air conditioned storage monitoring system |
| CN108898218A (en) * | 2018-05-24 | 2018-11-27 | 阿里巴巴集团控股有限公司 | A kind of training method of neural network model, device and computer equipment |
| CN109816150A (en) * | 2018-12-29 | 2019-05-28 | 中国农业大学 | One kind adopting rear Table Grape forecasting shelf life method and device |
| WO2020140377A1 (en) * | 2019-01-04 | 2020-07-09 | 平安科技（深圳）有限公司 | Neural network model training method and apparatus, computer device, and storage medium |
| CN109840854A (en) * | 2019-03-04 | 2019-06-04 | 中国科学技术大学 | Grain mildew prediction technique and device |
| CN110210680A (en) * | 2019-06-11 | 2019-09-06 | 北京农业信息技术研究中心 | A kind of fish body Noninvasive Measuring Method of Freshness and device based on temperature change |
| CN113433081A (en) * | 2021-07-06 | 2021-09-24 | 江南大学 | Dynamic neural network-based quality detection method and system in fruit and vegetable drying process |
| CN113791055A (en) * | 2021-08-17 | 2021-12-14 | 北京农业信息技术研究中心 | Fish freshness detection method and system |
Non-Patent Citations (1)
| Title |
|---|
| 张永超 等: "基于环境气体信息的BP神经网络苹果贮藏品质预测", 江苏农业学报, vol. 36, no. 1, pages 194 - 198 * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| do Nascimento Nunes et al. | Improvement in fresh fruit and vegetable logistics quality: berry logistics field studies | |
| AU2016286280B2 (en) | Combined method for detecting anomalies in a water distribution system | |
| CN111222549B (en) | Unmanned aerial vehicle fault prediction method based on deep neural network | |
| Melesse et al. | Machine learning-based digital twin for monitoring fruit quality evolution | |
| CN110210680A (en) | A kind of fish body Noninvasive Measuring Method of Freshness and device based on temperature change | |
| Zarifneshat et al. | Predictions of apple bruise volume using artificial neural network | |
| Sciortino et al. | A webGIS-based system for real time shelf life prediction | |
| US11748766B2 (en) | System and method for robust, modular, product sensitive monitoring and encoding of quality and safety along the perishable supply chain | |
| CN112070357A (en) | Radar radiation source threat assessment method based on improved BP neural network | |
| CN113327172A (en) | Grain condition data outlier detection method based on isolated forest | |
| Lin et al. | Survey on operational perishables quality control and logistics | |
| CN115049026A (en) | Regression analysis method of space non-stationarity relation based on GSNNR | |
| Koranga et al. | Analysis of white wine using machine learning algorithms | |
| Brezmes et al. | Electronic noses for monitoring the quality of fruit | |
| Jedermann et al. | Linking keeping quality models and sensor systems to an autonomous transport supervision system | |
| CN115184395A (en) | Fruit and vegetable weight loss rate prediction method and device, electronic equipment and storage medium | |
| CN116542565B (en) | Pet puffed food management method and system based on proportion detection technology | |
| CN116972913A (en) | On-line monitoring method and system for running state of cold chain equipment | |
| Libertino et al. | Number of consumers necessary for survival analysis estimations based on each consumer evaluating a single sample | |
| Rajini et al. | Selection of significant features and prediction of red winequality using logistic regression | |
| Khoo et al. | Kiwi fruit IoT shelf life estimation during transportation with cloud computing | |
| CN115009744B (en) | Method and device for predicting delivery duration of front cargo | |
| US12033165B2 (en) | System and method for perishable product monitoring along a supply chain | |
| Berruto et al. | Hybrid modelling for fruit quality prediction in supply chain networks | |
| Nwaizu et al. | Application of digital twin in evaluating quality changes in tomato value-chain in Nigeria |
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 |