Method for predicting comprehensive quality of blueberry fruit

Abstract

The present invention discloses a method for predicting the comprehensive quality of blueberry fruit, which relates to the field of blueberry fruit quality prediction. The method comprises the following steps: extracting principal component characteristics of blueberry fruit at each stage and establishing a data standard for principal component characteristics of blueberry fruit at each stage; establishing a blueberry fruit quality balance coefficient and obtaining the blueberry fruit quality balance coefficient value of blueberry fruit in each sampling period; establishing an influence weight of the principal component characteristics of blueberry fruit and obtaining the influence of different principal component characteristics on the quality of blueberry fruit; establishing a blueberry fruit optimization scheme; establishing a blueberry fruit comprehensive quality prediction model and improving the accuracy and reliability of the model prediction through a feedback neural network; establishing a blueberry fruit comprehensive quality prediction platform and correcting and optimizing the model by inputting optimization parameters. The method effectively predicts the comprehensive quality of blueberry fruit based on the principal component characteristic data of the blueberry fruit growth stage, and ensures the accuracy and reliability of the prediction result through feedback.

Description

Technical Field

[0001] The present invention relates to the field of blueberry fruit quality prediction, and in particular to a method for predicting the comprehensive quality of blueberry fruits. Background Art

[0002] The growth of blueberry fruit is divided into five stages: fruit setting, swelling, color change, maturity and over-maturity. The appearance, color and main components of blueberry fruit show obvious changes in each stage. By analyzing the appearance, color and main components of blueberry fruit in each stage, the comprehensive quality of blueberry fruit can be effectively predicted. Based on the comprehensive prediction results of blueberry fruit quality, the quality of blueberry fruit can be improved by artificial interference, thereby greatly improving the comprehensive quality of blueberry fruit, reducing the inferior quality rate of blueberry fruit, and effectively improving the production efficiency and yield rate of blueberry fruit.

[0003] The existing technology mainly focuses on the analysis and comprehensive evaluation of blueberry fruit quality. It is a technology based on the evaluation of blueberry fruit at a mature or single stage. It lacks analysis of the overall growth stage of the blueberry fruit, resulting in inaccurate prediction results of the comprehensive quality of the blueberry fruit, or the prediction results of the comprehensive quality of the blueberry fruit have a lag, which easily leads to waste of resources and is not conducive to improving the production efficiency and yield rate of blueberry fruit. Summary of the Invention

[0004] In order to solve the above technical problems, a method for predicting the comprehensive quality of blueberry fruits is provided. This technical solution solves the problem raised in the above background technology that there is a lack of analysis of the overall growth stage of blueberry fruits, which leads to inaccurate prediction results of the comprehensive quality of blueberry fruits, or the prediction results of the comprehensive quality of blueberry fruits have a lag, which easily leads to waste of resources and is not conducive to improving the production efficiency and yield rate of blueberry fruits.

[0005] In order to achieve the above objects, the technical solution adopted by the present invention is:

[0006] A method for predicting the comprehensive quality of blueberry fruit, comprising:

[0007] Based on big data, the principal component characteristics of blueberry fruit at each stage are extracted, and the data standard of principal component characteristics of blueberry fruit at each stage is established;

[0008] Setting a sampling period, establishing a blueberry fruit quality balance coefficient, and obtaining the blueberry fruit quality balance coefficient value of the blueberry fruit in each sampling period;

[0009] According to the main component characteristics of blueberry fruit, the influence weights of the main component characteristics of blueberry fruit were established to obtain the effects of different main component characteristics on the quality of blueberry fruit.

[0010] Based on big data, we can obtain the characteristics of the external environment, trace elements, and soil quality that affect the main components of blueberry fruit, and establish a blueberry fruit optimization plan;

[0011] Based on the neural network algorithm, a comprehensive quality prediction model for blueberry fruit was established, and the accuracy and reliability of the model prediction were improved through the feedback neural network.

[0012] A comprehensive blueberry fruit quality prediction platform was established, and the model was corrected and optimized by analyzing and storing periodic blueberry fruit quality data and inputting optimization parameters.

[0013] Preferably, the extraction of principal component characteristics of blueberry fruits at each stage based on big data and the establishment of a data standard for principal component characteristics of blueberry fruits at each stage specifically include:

[0014] According to the growth of blueberry fruit, the growth of blueberry fruit is divided into five stages: fruit setting stage, swelling stage, color change stage, maturity stage and over-maturity stage;

[0015] Based on the blueberry quality assessment method, the changes in appearance, color, and main components of blueberry fruits in these five stages were extracted. The main component characteristics include: appearance, color, and main components. The main components include: sugar-acid ratio, anthocyanins, vitamins, total phenols, and volatile flavor substances.

[0016] Based on big data, the change data of the principal component characteristics of blueberry fruits in these five stages are obtained, and according to the change data of the principal component characteristics, the data standards of the principal component characteristics of blueberry fruits at different stages are set.

[0017] Preferably, setting the sampling period, establishing the blueberry fruit quality balance coefficient, and obtaining the blueberry fruit quality balance coefficient value of the blueberry fruit in each sampling period specifically includes:

[0018] According to the growth stage of blueberry fruit, a blueberry fruit detection sampling period is set in each stage to obtain the periodic growth data of blueberry fruit in each stage;

[0019] Optimize the normalization and filtering of the periodic growth data of blueberry fruits at each stage;

[0020] Based on the optimized periodic growth data of blueberry fruits at each stage, a periodic growth dataset of blueberry fruits at each stage is established;

[0021] Based on the data sets of blueberry fruit growth at each stage, the blueberry fruit quality balance coefficient is established, and the blueberry fruit quality balance coefficient value of blueberry fruit in each sampling period is obtained;

[0022] The blueberry fruit quality balance coefficient expression is:

[0023]

[0024] Where, For the Phase I The quality balance coefficient of blueberry fruit sampled by For the The weight of the growth stage on blueberry fruit quality, For the The principal component features correspond to the The principal component feature detection data of blueberry fruits, For the The standard data of blueberry fruit principal component characteristics corresponding to the principal component characteristics, For the The number of blueberry fruit samples collected, is the equilibrium constant term.

[0025] Preferably, establishing the influence weights of the principal component characteristics of blueberry fruit based on the principal component characteristics of blueberry fruit and obtaining the influence of different principal component characteristics on the quality of blueberry fruit specifically includes:

[0026] Based on big data, we obtain evaluation values ​​for blueberries of the same type but different qualities, and obtain a principal component feature data cluster for each evaluation value. A principal component feature data cluster refers to a set of multiple principal component feature data that simultaneously meet a single blueberry fruit evaluation value.

[0027] According to the evaluation values ​​of blueberry fruits of the same type and different qualities, a set of blueberry fruit quality evaluation values ​​is established;

[0028] According to the blueberry fruit principal component characteristic data cluster, a set of blueberry fruit different principal component characteristic data clusters is established;

[0029] Based on the blueberry fruit quality evaluation value set and the blueberry fruit different principal component characteristic data cluster set, the influence weight of the blueberry fruit principal component characteristics is established;

[0030] According to the influence weights of the main component characteristics of blueberry fruit, a data set of the influence weights of different main component characteristics of blueberry fruit is established;

[0031] The influence weight expression of the main component characteristics of blueberry fruit is:

[0032]

[0033] Where, For the The correlation between the principal component characteristics and the blueberry fruit quality evaluation value, that is, the first The influence weight of the principal component features, For the The blueberry fruit quality evaluation value corresponds to The number of groups in the principal component feature data cluster, is the number of numerical types for blueberry fruit quality evaluation, For the The principal component features correspond to the The collected data in the data cluster set, For the The principal component features correspond to the The target data in the data cluster set, For the Blueberry fruit quality evaluation value, Numerical target value for blueberry fruit quality evaluation.

[0034] Preferably, the method of obtaining the characteristics of the external environment, trace elements and soil quality that affect the main component characteristics of blueberry fruit based on big data and establishing a blueberry fruit optimization plan specifically includes:

[0035] Based on big data, we obtained the characteristics of the external environment, trace elements, and soil quality that affect the main component characteristics of blueberry fruit. The external environmental characteristics include temperature, humidity, and light intensity, and the soil quality characteristics include pH, soil density, and nutrient content.

[0036] Establish the corresponding relationship between the main component characteristics of blueberry fruit and the external environment, trace elements and soil quality characteristics, and set the corresponding blueberry fruit optimization plan based on the relationship between the two;

[0037] Record and quantify the input optimization amount in the blueberry fruit optimization plan, and set the input optimization amount as the input optimization parameter;

[0038] Based on the implementation results of each blueberry fruit optimization plan, a set of blueberry fruit input optimization parameters is established.

[0039] Preferably, the method of establishing a blueberry fruit comprehensive quality prediction model based on a neural network algorithm and improving the accuracy and reliability of the model prediction through a feedback neural network specifically includes:

[0040] According to the blueberry fruit quality evaluation value set, the blueberry fruit quality evaluation value set is used as target output data;

[0041] The blueberry fruit quality balance coefficient and the influence weight of the main component characteristics of blueberry fruit in each collection period are used as the input data of the blueberry fruit comprehensive quality prediction model;

[0042] According to the blueberry fruit input optimization parameter set, the blueberry fruit input optimization parameter set is used as input data of the feedback layer;

[0043] Construct a feedback function in the feedback layer, establish a comprehensive blueberry fruit quality prediction model, and improve the accuracy and reliability of the model prediction through a feedback neural network;

[0044] Based on the blueberry fruit comprehensive quality prediction model, the blueberry fruit quality is comprehensively predicted and evaluated, and the prediction results are output;

[0045] The feedback function expression of the feedback layer is:

[0046]

[0047] Where, is the input data of the blueberry fruit comprehensive quality prediction model at the current moment, is the input data of the blueberry fruit comprehensive quality prediction model at the previous moment, is the input data of the feedback layer at the current moment, is the output data of the blueberry fruit comprehensive quality prediction model at the previous moment, The adaptation function for the input data of the blueberry fruit comprehensive quality prediction model, is the feedback function.

[0048] Preferably, the establishment of a blueberry fruit comprehensive quality prediction platform, by analyzing and storing periodic blueberry fruit quality data, and by inputting optimization parameters, the model is corrected and optimized, specifically comprising:

[0049] Establish a blueberry fruit comprehensive quality prediction platform to build an operating environment for the blueberry fruit comprehensive quality prediction model and ensure the normal operation of the model;

[0050] Based on the blueberry fruit comprehensive quality prediction platform, it is used to receive and store the principal component feature data collected periodically at each stage;

[0051] Based on the blueberry fruit comprehensive quality prediction platform, the model is modified and optimized by analyzing and storing periodic blueberry fruit quality data and inputting optimization parameters;

[0052] Based on the comprehensive blueberry fruit quality prediction platform, the corresponding blueberry fruit optimization plan is obtained according to the characteristic feedback of the external environment, trace elements and soil quality.

[0053] Compared with the prior art, the present invention has the following beneficial effects:

[0054] According to the growth of blueberry fruit, the growth of blueberry fruit is divided into five stages: fruit setting stage, swelling stage, color change stage, maturity stage and over-maturity stage. By extracting the change characteristics of the appearance, color and main components of blueberry fruit in these five stages, the stage-by-stage blueberry fruit principal component characteristic data standard is set. Secondly, according to the growth stage of blueberry fruit, the blueberry fruit detection sampling cycle is set in each stage to obtain the periodic growth data of blueberry fruit in each stage, and the blueberry fruit quality balance coefficient is established based on the data to obtain the blueberry fruit quality balance coefficient value of blueberry fruit in each sampling period. Furthermore, based on big data, the evaluation values ​​of blueberry fruits of the same type but different qualities and the blueberry fruit principal component characteristic data cluster under each evaluation value are obtained to establish the blueberry fruit principal component characteristic influence. The effect weights are calculated to obtain the influence of different principal component characteristics on the quality of blueberry fruit. According to the characteristics of the external environment, trace elements and soil quality that affect the principal component characteristics of blueberry fruit, the corresponding relationship between the principal component characteristics of blueberry fruit and the external environment, trace elements and soil quality characteristics is established. At the same time, according to the corresponding relationship between the two, a blueberry fruit optimization plan is established. Finally, based on the neural network algorithm, a feedback function of the feedback layer is constructed, and a blueberry fruit comprehensive quality prediction model is established. The blueberry fruit quality is comprehensively predicted and evaluated through the feedback neural network, and the prediction results are output, so that the comprehensive quality of blueberry fruit is effectively predicted according to the principal component characteristic data of the blueberry fruit growth stage, and the accuracy and reliability of the prediction results are guaranteed through feedback. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] Figure 1 This is a flow chart of the method for predicting the comprehensive quality of blueberry fruit of the present invention;

[0056] Figure 2 A flow chart of setting sampling periods, establishing a blueberry fruit quality balance coefficient, and obtaining a blueberry fruit quality balance coefficient value for blueberry fruits in each sampling period according to the present invention;

[0057] Figure 3 The present invention is a flow chart for establishing the influence weights of the main component characteristics of blueberry fruit based on the main component characteristics of blueberry fruit and obtaining the influence of different main component characteristics on the quality of blueberry fruit;

[0058] Figure 4 The invention uses big data to obtain the characteristics of the external environment, trace elements, and soil quality that affect the main component characteristics of blueberry fruit, and establishes a flow chart for the blueberry fruit optimization program;

[0059] Figure 5 The present invention is based on the neural network algorithm to establish a blueberry fruit comprehensive quality prediction model, and improve the accuracy and reliability of the model prediction flow chart by using a feedback neural network;

[0060] Figure 6This is a structural diagram of the electronic device proposed by the present invention;

[0061] Figure 7 This is a schematic diagram of the structure of the computer-readable storage medium proposed in the present invention. DETAILED DESCRIPTION

[0062] The following description is intended to disclose the present invention so that those skilled in the art can implement the present invention. The preferred embodiments described below are only examples, and those skilled in the art may conceive of other obvious variations.

[0063] Reference Figure 1 As shown, a method for predicting the comprehensive quality of blueberry fruit comprises:

[0064] Based on big data, the principal component characteristics of blueberry fruit at each stage are extracted, and the data standard of principal component characteristics of blueberry fruit at each stage is established;

[0065] Setting a sampling period, establishing a blueberry fruit quality balance coefficient, and obtaining the blueberry fruit quality balance coefficient value of the blueberry fruit in each sampling period;

[0066] According to the main component characteristics of blueberry fruit, the influence weights of the main component characteristics of blueberry fruit were established to obtain the effects of different main component characteristics on the quality of blueberry fruit.

[0067] Based on big data, we can obtain the characteristics of the external environment, trace elements, and soil quality that affect the main components of blueberry fruit, and establish a blueberry fruit optimization plan;

[0068] Based on the neural network algorithm, a comprehensive quality prediction model for blueberry fruit was established, and the accuracy and reliability of the model prediction were improved through the feedback neural network.

[0069] A comprehensive blueberry fruit quality prediction platform was established, and the model was corrected and optimized by analyzing and storing periodic blueberry fruit quality data and inputting optimization parameters.

[0070] It can be explained that this scheme divides the growth of blueberry fruit into five stages: fruit setting stage, swelling stage, color change stage, maturity stage and over-maturity stage according to the growth conditions of blueberry fruit. By extracting the change characteristics of the appearance, color and main components of blueberry fruit in these five stages, the stage-by-stage blueberry fruit principal component characteristic data standard is set. Secondly, according to the growth stage of blueberry fruit, a blueberry fruit detection sampling cycle is set in each stage to obtain the periodic growth data of blueberry fruit in each stage, and the blueberry fruit quality balance coefficient is established based on the data to obtain the blueberry fruit quality balance coefficient value of blueberry fruit in each sampling period. Furthermore, based on big data, the evaluation values ​​of blueberry fruits of the same type but different qualities and the blueberry fruit principal component characteristic data cluster under each evaluation value are obtained to establish the blueberry fruit principal component characteristic data cluster. The influence weights of component characteristics are calculated to obtain the influence of different principal component characteristics on the quality of blueberry fruit. According to the characteristics of the external environment, trace elements and soil quality that affect the principal component characteristics of blueberry fruit, the corresponding relationship between the principal component characteristics of blueberry fruit and the external environment, trace elements and soil quality characteristics is established. At the same time, according to the corresponding relationship between the two, a blueberry fruit optimization plan is established. Finally, based on the neural network algorithm, the feedback function of the feedback layer is constructed, and a blueberry fruit comprehensive quality prediction model is established. The blueberry fruit quality is comprehensively predicted and evaluated through the feedback neural network, and the prediction results are output, so as to effectively predict the comprehensive quality of blueberry fruit according to the principal component characteristic data of the blueberry fruit growth stage, and the accuracy and reliability of the prediction results are guaranteed through feedback.

[0071] Reference Figure 2 As shown, the setting of the sampling period, establishing the blueberry fruit quality balance coefficient, and obtaining the blueberry fruit quality balance coefficient value of the blueberry fruit in each sampling period specifically include:

[0072] According to the growth stage of blueberry fruit, a blueberry fruit detection sampling period is set in each stage to obtain the periodic growth data of blueberry fruit in each stage;

[0073] Optimize the normalization and filtering of the periodic growth data of blueberry fruits at each stage;

[0074] Based on the optimized periodic growth data of blueberry fruits at each stage, a periodic growth dataset of blueberry fruits at each stage is established;

[0075] Based on the data sets of blueberry fruit growth at each stage, the blueberry fruit quality balance coefficient is established, and the blueberry fruit quality balance coefficient value of blueberry fruit in each sampling period is obtained;

[0076] The blueberry fruit quality balance coefficient expression is:

[0077]

[0078] Where, For the Phase I The quality balance coefficient of blueberry fruit sampled by For the The weight of the growth stage on blueberry fruit quality, For the The principal component features correspond to the The principal component feature detection data of blueberry fruits, For the The standard data of blueberry fruit principal component characteristics corresponding to the principal component characteristics, For the The number of blueberry fruit samples collected, is the equilibrium constant term.

[0079] It can be explained that the growth of blueberry fruit is divided into five stages: fruit setting, swelling, color change, maturity and over-maturity. The appearance, color and main component characteristics of blueberry fruit in each stage have obvious changes. By analyzing the appearance, color and main component characteristic data of blueberry fruit in each stage, the comprehensive quality of blueberry fruit can be effectively predicted. Therefore, through the comprehensive prediction results of blueberry fruit quality, the quality of blueberry fruit can be improved by artificial interference. Therefore, it is necessary to set a blueberry fruit detection sampling cycle for the blueberry fruit growth stage, and timely discover the abnormal growth of blueberry fruit through the periodic growth data of each stage. This scheme obtains the blueberry fruit quality balance coefficient value of blueberry fruit in each sampling period, and performs a balanced analysis on the main component characteristic data of blueberry fruit in each stage to reduce data redundancy.

[0080] Reference Figure 3 As shown, the method of establishing the influence weight of the main component characteristics of blueberry fruit based on the main component characteristics of blueberry fruit and obtaining the influence of different main component characteristics on the quality of blueberry fruit specifically includes:

[0081] Based on big data, we obtain evaluation values ​​for blueberries of the same type but different qualities, and obtain a principal component feature data cluster for each evaluation value. A principal component feature data cluster refers to a set of multiple principal component feature data that simultaneously meet a single blueberry fruit evaluation value.

[0082] According to the evaluation values ​​of blueberry fruits of the same type and different qualities, a set of blueberry fruit quality evaluation values ​​is established;

[0083] According to the blueberry fruit principal component characteristic data cluster, a set of blueberry fruit different principal component characteristic data clusters is established;

[0084] Based on the blueberry fruit quality evaluation value set and the blueberry fruit different principal component characteristic data cluster set, the influence weight of the blueberry fruit principal component characteristics is established;

[0085] According to the influence weights of the main component characteristics of blueberry fruit, a data set of the influence weights of different main component characteristics of blueberry fruit is established;

[0086] The influence weight expression of the main component characteristics of blueberry fruit is:

[0087]

[0088] Where, For the The correlation between the principal component characteristics and the blueberry fruit quality evaluation value, that is, the first The influence weight of the principal component features, For the The blueberry fruit quality evaluation value corresponds to The number of groups in the principal component feature data cluster, is the number of numerical types for blueberry fruit quality evaluation, For the The principal component features correspond to the The collected data in the data cluster set, For the The principal component features correspond to the The target data in the data cluster set, For the Blueberry fruit quality evaluation value, Numerical target value for blueberry fruit quality evaluation.

[0089] It can be explained that different principal component characteristics of blueberry fruit have different effects on the final results of blueberry fruit quality evaluation. Therefore, when comprehensively evaluating the quality of blueberry fruit, it is necessary to give priority to the impact of different principal component characteristics on the quality of blueberry fruit. This scheme establishes the influence weights of the principal component characteristics of blueberry fruit and analyzes the correlation between the different principal component characteristics of blueberry fruit and the blueberry fruit quality evaluation values, so as to effectively judge the degree of influence of different principal component characteristics of blueberry fruit on the blueberry fruit quality evaluation values.

[0090] Reference Figure 4 As shown, the method of obtaining the characteristics of the external environment, trace elements and soil quality that affect the main component characteristics of blueberry fruit based on big data and establishing a blueberry fruit optimization plan specifically includes:

[0091] Based on big data, we obtained the characteristics of the external environment, trace elements, and soil quality that affect the main component characteristics of blueberry fruit. The external environmental characteristics include temperature, humidity, and light intensity, and the soil quality characteristics include pH, soil density, and nutrient content.

[0092] Establish the corresponding relationship between the main component characteristics of blueberry fruit and the external environment, trace elements and soil quality characteristics, and set the corresponding blueberry fruit optimization plan based on the relationship between the two;

[0093] Record and quantify the input optimization amount in the blueberry fruit optimization plan, and set the input optimization amount as the input optimization parameter;

[0094] Based on the implementation results of each blueberry fruit optimization plan, a set of blueberry fruit input optimization parameters is established.

[0095] What can be explained is that by analyzing the characteristics of the external environment, trace elements and soil quality that affect the main component characteristics of blueberry fruit, establishing the corresponding relationship between the main component characteristics of blueberry fruit and the external environment, trace elements and soil quality characteristics, and taking corresponding blueberry fruit optimization plans based on the relationship between the two, the quality of blueberry fruit can be effectively improved through human intervention, which greatly improves the overall quality of blueberry fruit, reduces the inferior quality rate of blueberry fruit, and effectively improves the production efficiency and yield rate of blueberry fruit. Among them, the corresponding relationship between the main component characteristics of blueberry fruit and the external environment, trace elements and soil quality characteristics can be explained. It can be explained as follows: for example, the content of the main component of anthocyanins increases during the color change period of blueberry fruit. By comparing the standard value of anthocyanin content in this stage, it can be effectively discovered whether the current anthocyanin content meets the standard. If it does not meet the standard, the content of anthocyanins in blueberry fruit can be increased by increasing light intensity, controlling humidity and supplementing fruit growth regulators. Therefore, by detecting the absolute difference between the anthocyanin value and the standard value, the optimization intensity of blueberry fruit can be determined, that is, the improvement intensity of the external environment, trace elements and soil quality characteristics, thereby establishing a correspondence between the main component characteristics of blueberry fruit and the external environment, trace elements and soil quality characteristics.

[0096] Reference Figure 5 As shown, the blueberry fruit comprehensive quality prediction model is established based on the neural network algorithm, and the accuracy and reliability of the model prediction are improved by the feedback neural network. Specifically, the following steps are involved:

[0097] According to the blueberry fruit quality evaluation value set, the blueberry fruit quality evaluation value set is used as target output data;

[0098] The blueberry fruit quality balance coefficient and the influence weight of the main component characteristics of blueberry fruit in each collection period are used as the input data of the blueberry fruit comprehensive quality prediction model;

[0099] According to the blueberry fruit input optimization parameter set, the blueberry fruit input optimization parameter set is used as input data of the feedback layer;

[0100] Construct a feedback function in the feedback layer, establish a comprehensive blueberry fruit quality prediction model, and improve the accuracy and reliability of the model prediction through a feedback neural network;

[0101] Based on the blueberry fruit comprehensive quality prediction model, the blueberry fruit quality is comprehensively predicted and evaluated, and the prediction results are output;

[0102] The feedback function expression of the feedback layer is:

[0103]

[0104] Where, is the input data of the blueberry fruit comprehensive quality prediction model at the current moment, is the input data of the blueberry fruit comprehensive quality prediction model at the previous moment, is the input data of the feedback layer at the current moment, is the output data of the blueberry fruit comprehensive quality prediction model at the previous moment, The adaptation function for the input data of the blueberry fruit comprehensive quality prediction model, is the feedback function.

[0105] It can be explained that when predicting the comprehensive quality of blueberry fruit, conventional neural network models are difficult to achieve the prediction results of feedback human intervention. Therefore, this scheme introduces a feedback layer into the hidden layer of the conventional neural network model by constructing a feedback function of the feedback layer. By combining the system output with the blueberry fruit input optimization parameters and re-injecting them into the input or intermediate layer, a closed loop is formed, thereby correcting the neural network model by combining the blueberry fruit input optimization parameters, and improving the accuracy and reliability of the model prediction through the feedback neural network. Among them, the prediction steps of the blueberry fruit comprehensive quality prediction model are:

[0106] Step 1: Based on the obtained blueberry fruit comprehensive quality prediction model input data, a preliminary prediction of the comprehensive quality of the blueberry fruit is made, and the deficiencies of the current principal component characteristic data of the blueberry fruit are obtained;

[0107] Step 2: adopting a blueberry fruit optimization plan, obtaining blueberry fruit input optimization parameters, and modifying the model by inputting the blueberry fruit input optimization parameters into the blueberry fruit comprehensive quality prediction model;

[0108] Step 3: Re-predict the comprehensive quality of blueberry fruit according to the feedback update of the blueberry fruit comprehensive quality prediction model, and repeat the operation until the optimal comprehensive quality of blueberry fruit is obtained;

[0109] Step 4: Based on the periodic principal component feature data analysis collected in stages, the accuracy of the model prediction is judged by comparing the prediction results of the blueberry fruit comprehensive quality prediction model with the actual test results, and the model is optimized by updating parameters or numerical optimization.

[0110] Furthermore, the method according to the embodiment of the present application can also be used with the aid of Figure 6 The electronic device architecture shown in FIG. Figure 6 As shown, the electronic device 500 may include a bus 501, one or more CPUs 502, a read-only memory (ROM) 503, a random access memory (RAM) 504, a communication port 505 connected to a network, an input / output component 506, a hard disk 507, etc. The storage device in the electronic device 500, such as the ROM 503 or the hard disk 507, may store a method for predicting the comprehensive quality of blueberry fruit provided in the present application. The electronic device 500 may also include a user interface 508. Of course, Figure 6 The architecture shown is only exemplary and can be omitted according to actual needs when implementing different devices. Figure 6 One or more components of an electronic device are shown.

[0111] Figure 7 This is a schematic diagram of the computer-readable storage medium structure provided by an embodiment of the present application. Figure 7 1 shows a computer-readable storage medium 600 according to one embodiment of the present application. Computer-readable instructions are stored on the computer-readable storage medium 600. When the computer-readable instructions are executed by a processor, the method for predicting the comprehensive quality of blueberry fruit according to an embodiment of the present application described with reference to the above figures can be executed. The storage medium 600 includes, but is not limited to, volatile memory and / or non-volatile memory. Volatile memory may include, for example, random access memory (RAM) and cache memory. Non-volatile memory may include, for example, read-only memory (ROM), a hard disk, flash memory, etc.

[0112] In summary, the advantages of the present invention are: effectively predicting the comprehensive quality of blueberry fruit based on the principal component characteristic data of the blueberry fruit growth stage, and ensuring the accuracy and reliability of the prediction results through feedback.

[0113] The above shows and describes the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The above embodiments and descriptions merely illustrate the principles of the present invention. Various changes and modifications may be made to the present invention without departing from the spirit and scope of the present invention. Such changes and modifications are intended to fall within the scope of the present invention. The scope of protection claimed by the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for predicting the comprehensive quality of blueberry fruit, characterized in that: include: Based on big data, the principal component characteristics of blueberry fruit at each stage are extracted, and the data standard of principal component characteristics of blueberry fruit at each stage is established; Setting a sampling period, establishing a blueberry fruit quality balance coefficient, and obtaining the blueberry fruit quality balance coefficient value of the blueberry fruit in each sampling period; According to the main component characteristics of blueberry fruit, the influence weights of the main component characteristics of blueberry fruit were established to obtain the effects of different main component characteristics on the quality of blueberry fruit. Based on big data, we can obtain the characteristics of the external environment, trace elements, and soil quality that affect the main components of blueberry fruit, and establish a blueberry fruit optimization plan; Based on the neural network algorithm, a comprehensive quality prediction model for blueberry fruit was established, and the accuracy and reliability of the model prediction were improved through the feedback neural network. Establish a comprehensive blueberry fruit quality prediction platform, analyze and store periodic blueberry fruit quality data, and input optimization parameters to revise and optimize the model; The steps of setting the sampling period, establishing the blueberry fruit quality balance coefficient, and obtaining the blueberry fruit quality balance coefficient value of the blueberry fruit in each sampling period specifically include: Based on the data sets of blueberry fruit growth at each stage, the blueberry fruit quality balance coefficient is established, and the blueberry fruit quality balance coefficient value of blueberry fruit in each sampling period is obtained; The blueberry fruit quality balance coefficient expression is: Where, is the blueberry fruit quality balance coefficient of the zth sampling in the cth stage, ε c is the weight of stage c on blueberry fruit quality growth, is the i-th blueberry fruit principal component feature detection data corresponding to the j-th principal component feature, is the standard data of the principal component characteristics of blueberry fruit corresponding to the jth principal component characteristic, n is the number of samples of blueberry fruit in the zth sampling, and q is the equilibrium constant term; The method of establishing the influence weight of the principal component characteristics of blueberry fruit based on the principal component characteristics of blueberry fruit and obtaining the influence of different principal component characteristics on the quality of blueberry fruit specifically includes: Based on the blueberry fruit quality evaluation value set and the blueberry fruit different principal component characteristic data cluster set, the influence weight of the blueberry fruit principal component characteristics is established; The influence weight expression of the main component characteristics of blueberry fruit is: Where r j is the correlation between the jth principal component feature and the blueberry fruit quality evaluation value, that is, the influence weight of the jth principal component feature of the blueberry fruit, s is the number of groups in the data cluster of the jth principal component feature corresponding to the i-th blueberry fruit quality evaluation value, m is the number of blueberry fruit quality evaluation value types, The jth principal component feature corresponds to the collected data in the i-th data cluster set, The jth principal component feature corresponds to the target data in the i-th data cluster set, b i is the quality evaluation value of the i-th blueberry fruit, Numerical target values ​​for blueberry fruit quality evaluation; The above-mentioned method of establishing a comprehensive blueberry fruit quality prediction model based on a neural network algorithm and improving the accuracy and reliability of the model prediction through a feedback neural network specifically includes: According to the blueberry fruit input optimization parameter set, the blueberry fruit input optimization parameter set is used as input data of the feedback layer; Construct a feedback function in the feedback layer, establish a comprehensive blueberry fruit quality prediction model, and improve the accuracy and reliability of the model prediction through a feedback neural network; The feedback function expression of the feedback layer is: G t =f(G t-1 ,u t )+g(h t-1 ) Where G t is the input data of the blueberry fruit comprehensive quality prediction model at the current moment, G t-1 is the input data of the blueberry fruit comprehensive quality prediction model at the previous moment, u t is the input data of the feedback layer at the current moment, h t-1 is the output data of the blueberry fruit comprehensive quality prediction model at the previous moment, f(G t-1 ,u t ) is the adaptation function of the input data of the blueberry fruit comprehensive quality prediction model, g(h t-1 ) is the feedback function.

2. The method for predicting the comprehensive quality of blueberry fruit according to claim 1, wherein: The method of extracting the principal component characteristics of blueberry fruits at each stage based on big data and establishing the principal component characteristic data standard of blueberry fruits at each stage specifically includes: According to the growth of blueberry fruit, the growth of blueberry fruit is divided into five stages: fruit setting stage, swelling stage, color change stage, maturity stage and over-maturity stage; Based on the blueberry quality assessment method, the changes in appearance, color, and main components of blueberry fruits in these five stages were extracted. The main component characteristics include: appearance, color, and main components. The main components include: sugar-acid ratio, anthocyanins, vitamins, total phenols, and volatile flavor substances. Based on big data, the change data of the principal component characteristics of blueberry fruits in these five stages are obtained, and according to the change data of the principal component characteristics, the data standards of the principal component characteristics of blueberry fruits at different stages are set.

3. The method for predicting the comprehensive quality of blueberry fruit according to claim 2, wherein: The steps of setting the sampling period, establishing the blueberry fruit quality balance coefficient, and obtaining the blueberry fruit quality balance coefficient value of the blueberry fruit in each sampling period specifically include: According to the growth stage of blueberry fruit, a blueberry fruit detection sampling period is set in each stage to obtain the periodic growth data of blueberry fruit in each stage; Optimize the normalization and filtering of the periodic growth data of blueberry fruits at each stage; Based on the optimized periodic growth data of blueberry fruits at different stages, a periodic growth dataset of blueberry fruits at different stages was established.

4. The method for predicting the comprehensive quality of blueberry fruit according to claim 3, wherein: The method of establishing the influence weight of the principal component characteristics of blueberry fruit based on the principal component characteristics of blueberry fruit and obtaining the influence of different principal component characteristics on the quality of blueberry fruit specifically includes: Based on big data, we obtain evaluation values ​​for blueberries of the same type but different qualities, and obtain a principal component feature data cluster for each evaluation value. A principal component feature data cluster refers to a set of multiple principal component feature data that simultaneously meet a single blueberry fruit evaluation value. According to the evaluation values ​​of blueberry fruits of the same type and different qualities, a set of blueberry fruit quality evaluation values ​​is established; According to the blueberry fruit principal component characteristic data cluster, a set of blueberry fruit different principal component characteristic data clusters is established; According to the influence weights of the principal component characteristics of blueberry fruit, a data set of the influence weights of different principal component characteristics of blueberry fruit is established.

5. The method for predicting the comprehensive quality of blueberry fruit according to claim 4, wherein: The method of obtaining the characteristics of the external environment, trace elements, and soil quality that affect the main component characteristics of blueberry fruit based on big data and establishing a blueberry fruit optimization plan specifically includes: Based on big data, we obtained the characteristics of the external environment, trace elements, and soil quality that affect the main component characteristics of blueberry fruit. The external environmental characteristics include temperature, humidity, and light intensity, and the soil quality characteristics include pH, soil density, and nutrient content. Establish the corresponding relationship between the main component characteristics of blueberry fruit and the external environment, trace elements and soil quality characteristics, and set the corresponding blueberry fruit optimization plan based on the relationship between the two; Record and quantify the input optimization amount in the blueberry fruit optimization plan, and set the input optimization amount as the input optimization parameter; Based on the implementation results of each blueberry fruit optimization plan, a set of blueberry fruit input optimization parameters is established.

6. The method for predicting comprehensive quality of blueberry fruit according to claim 5, characterized in that: The above-mentioned method of establishing a comprehensive blueberry fruit quality prediction model based on a neural network algorithm and improving the accuracy and reliability of the model prediction through a feedback neural network specifically includes: According to the blueberry fruit quality evaluation value set, the blueberry fruit quality evaluation value set is used as target output data; The blueberry fruit quality balance coefficient and the influence weight of the main component characteristics of blueberry fruit in each collection period are used as the input data of the blueberry fruit comprehensive quality prediction model; Based on the comprehensive quality prediction model of blueberry fruit, the quality of blueberry fruit is comprehensively predicted and evaluated, and the prediction results are output.

7. The method for predicting comprehensive quality of blueberry fruit according to claim 6, characterized in that: The establishment of a blueberry fruit comprehensive quality prediction platform, by analyzing and storing periodic blueberry fruit quality data and inputting optimization parameters to modify and optimize the model specifically includes: Establish a blueberry fruit comprehensive quality prediction platform to build an operating environment for the blueberry fruit comprehensive quality prediction model and ensure the normal operation of the model; Based on the blueberry fruit comprehensive quality prediction platform, it is used to receive and store the principal component feature data collected periodically at each stage; Based on the blueberry fruit comprehensive quality prediction platform, the model is modified and optimized by analyzing and storing periodic blueberry fruit quality data and inputting optimization parameters; Based on the comprehensive blueberry fruit quality prediction platform, the corresponding blueberry fruit optimization plan is obtained according to the characteristic feedback of the external environment, trace elements and soil quality.

8. An electronic device, characterized in that: include: at least one processor; and a memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to execute the method for predicting the comprehensive quality of blueberry fruits as described in any one of claims 1 to 7.

9. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the method for predicting the comprehensive quality of blueberry fruit according to any one of claims 1 to 7 is implemented.

Images