CN117789095A - Cut flower opening period optimization method, system, equipment and storage medium - Google Patents
Cut flower opening period optimization method, system, equipment and storage medium Download PDFInfo
- Publication number
- CN117789095A CN117789095A CN202410002772.2A CN202410002772A CN117789095A CN 117789095 A CN117789095 A CN 117789095A CN 202410002772 A CN202410002772 A CN 202410002772A CN 117789095 A CN117789095 A CN 117789095A
- Authority
- CN
- China
- Prior art keywords
- flower
- parameter
- value
- parameters
- next moment
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 238000000034 method Methods 0.000 title claims abstract description 35
- 238000005457 optimization Methods 0.000 title claims abstract description 14
- 238000003860 storage Methods 0.000 title claims abstract description 9
- 238000012423 maintenance Methods 0.000 claims abstract description 65
- 238000013527 convolutional neural network Methods 0.000 claims abstract description 17
- 230000006870 function Effects 0.000 claims description 183
- 230000009471 action Effects 0.000 claims description 85
- 238000012544 monitoring process Methods 0.000 claims description 65
- 238000013139 quantization Methods 0.000 claims description 18
- 238000002372 labelling Methods 0.000 claims description 17
- 238000012549 training Methods 0.000 claims description 17
- 230000036541 health Effects 0.000 claims description 12
- 235000015097 nutrients Nutrition 0.000 claims description 10
- 238000005286 illumination Methods 0.000 claims description 9
- 238000012545 processing Methods 0.000 claims description 8
- 238000004364 calculation method Methods 0.000 claims description 7
- 238000000605 extraction Methods 0.000 claims description 6
- 238000005520 cutting process Methods 0.000 claims description 5
- 238000007781 pre-processing Methods 0.000 claims description 4
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 claims description 4
- 230000005284 excitation Effects 0.000 claims description 3
- 230000017260 vegetative to reproductive phase transition of meristem Effects 0.000 abstract description 6
- 230000002787 reinforcement Effects 0.000 abstract description 5
- 238000004458 analytical method Methods 0.000 abstract description 3
- 230000002035 prolonged effect Effects 0.000 abstract description 3
- 238000013473 artificial intelligence Methods 0.000 abstract description 2
- 230000000875 corresponding effect Effects 0.000 description 34
- 239000011159 matrix material Substances 0.000 description 11
- 238000013461 design Methods 0.000 description 6
- 238000010586 diagram Methods 0.000 description 3
- 238000004590 computer program Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 238000004836 empirical method Methods 0.000 description 2
- 230000002093 peripheral effect Effects 0.000 description 2
- 238000011176 pooling Methods 0.000 description 2
- 230000008569 process Effects 0.000 description 2
- 238000012360 testing method Methods 0.000 description 2
- 230000009466 transformation Effects 0.000 description 2
- PXFBZOLANLWPMH-UHFFFAOYSA-N 16-Epiaffinine Natural products C1C(C2=CC=CC=C2N2)=C2C(=O)CC2C(=CC)CN(C)C1C2CO PXFBZOLANLWPMH-UHFFFAOYSA-N 0.000 description 1
- ORILYTVJVMAKLC-UHFFFAOYSA-N Adamantane Natural products C1C(C2)CC3CC1CC2C3 ORILYTVJVMAKLC-UHFFFAOYSA-N 0.000 description 1
- 230000004913 activation Effects 0.000 description 1
- 238000003491 array Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000004422 calculation algorithm Methods 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000013136 deep learning model Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 239000013505 freshwater Substances 0.000 description 1
- 238000010438 heat treatment Methods 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 210000001503 joint Anatomy 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 238000010606 normalization Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 230000000737 periodic effect Effects 0.000 description 1
- 238000010008 shearing Methods 0.000 description 1
- 238000005728 strengthening Methods 0.000 description 1
- 230000009897 systematic effect Effects 0.000 description 1
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention belongs to the technical field of artificial intelligence, and particularly discloses a cut flower opening period optimization method, a system, equipment and a storage medium. The invention can realize automatic maintenance of cut flowers; the state characteristics of flowers can be automatically extracted through image recognition and analysis based on a convolutional neural network; based on the decision-making mode of reinforcement learning, the automatic intervention of the whole life cycle of the cut flowers can be realized, so that the flowering period of the cut flowers can be prolonged.
Description
Technical Field
The invention belongs to the technical field of artificial intelligence, and particularly relates to a cut flower opening period optimization method, a cut flower opening period optimization system, cut flower opening period optimization equipment and a storage medium.
Background
The flower arrangement art not only can improve the aesthetic feeling of the home, but also can increase the comfort level of living space, and is an important living part of many families. In a home setting, maintaining cut flowers bright and extending their open period is a common and often challenging task. After leaving the natural growth environment, the cut flowers gradually weaken their vitality, so that careful care and management are required to maintain their flowering phase.
Currently, most households rely on empirical methods to extend the life of the cut flowers, such as periodic replacement of fresh water, shearing out withered parts, and maintaining proper indoor temperature. Although effective, these methods often lack systematic and scientific basis and cannot be tailored to the specific needs of the various flowers. In particular, the limitations of the prior empirical methods are mainly manifested in the following aspects:
non-personalized treatment: home users often lack professional knowledge and cannot provide optimum fresh-keeping conditions for different types of cut flowers;
lack of precise control: the maintenance conditions of flowers, such as humidity, illumination and the like, are difficult to accurately control by ordinary families;
the maintenance is inconvenient: the manual maintenance of the flower arrangement is tedious and easy to forget, especially in busy life;
the information is insufficient: the real-time monitoring of the state of flowers is usually lacking, so that proper fresh-keeping measures cannot be taken in time.
Thus, there is a need for a method that can intelligently and automatically maintain and extend the flower arrangement open period in a household.
Disclosure of Invention
The invention aims to provide a method, a system, equipment and a storage medium for optimizing a cutting and opening cycle, which are used for solving the problems in the prior art.
In order to achieve the above purpose, the present invention adopts the following technical scheme:
in a first aspect, a method for optimizing a cut-flower opening cycle is provided, including:
a. acquiring a flower monitoring data set at the current moment, wherein the flower monitoring data set comprises a flower monitoring image and a plurality of growth environment monitoring parameters;
b. extracting features of the monitoring images to obtain a plurality of flower image feature parameters, taking each flower image feature parameter and each growth environment monitoring parameter as corresponding state parameters, and forming a state set at the current moment by utilizing each state parameter;
c. acquiring an action set and an initial value function value table, wherein the action set comprises a plurality of maintenance action parameters, the initial value function value table comprises a plurality of preset value function values, and each value function value is associated with a corresponding state parameter and action parameter;
d. calculating a set strategy function based on the initial value function value table and the state set at the current moment to obtain a corresponding strategy function value;
e. selecting action parameters of the next moment from the action set according to the strategy function values, and transmitting the action parameters of the next moment to the cut flower maintenance end, so that the cut flower maintenance end executes the action parameters of the next moment to carry out flower maintenance;
f. collecting a flower monitoring data set at the next moment, and executing the step b on the flower monitoring data set at the next moment to obtain a state set at the next moment;
g. calculating and determining a reward parameter of the next moment by adopting a preset reward function based on the state set and the action set of the next moment;
h. updating each value function value according to the state set and the rewarding parameter at the next moment to obtain an updated value function value table;
i. calculating a set strategy function based on the updated value function value table and a state set at the next moment to obtain a strategy function value corresponding to the next moment;
j. iteratively executing the steps e-i based on the strategy function value corresponding to the next moment until the set iteration condition is met, so as to obtain a final value function value table;
k. and deploying the strategy function and the final value function value table to the cut flower maintenance end so that the cut flower maintenance end performs cut flower open period maintenance by using the strategy function and the final value function value table.
In one possible design, the feature extraction of the monitored image to obtain a plurality of flower image feature parameters includes:
and inputting the monitoring image into a convolutional neural network model trained by a training set to extract image features to obtain a plurality of flower image feature parameters, wherein the training set comprises a plurality of flower image samples marked with corresponding image feature quantization indexes.
In one possible design, before feature extraction of the monitored image, the method further includes:
collecting a plurality of flower images;
corresponding image characteristic quantization index labeling is carried out on each flower image, wherein the image characteristic quantization index labeling comprises flower openness quantization index labeling, color vividness quantization index labeling and health state quantization index labeling, and each flower image after labeling is utilized to form an initial image set;
performing image preprocessing and image enhancement processing on each flower image in the initial image set to obtain a training set;
training a convolutional neural network model constructed in advance by using a training set to obtain a trained convolutional neural network model.
In one possible design, the number of growth environment monitoring parameters includes an indoor temperature parameter, a relative humidity parameter, and an illumination intensity parameter; the flower image characteristic parameters comprise flower opening degree parameters, color vividness parameters and health state parameters; the plurality of maintenance action parameters comprise a water supply adjustment parameter, a nutrient solution supply adjustment parameter, an illumination intensity adjustment parameter, an indoor temperature adjustment parameter and an indoor humidity adjustment parameter.
In one possible design, the policy function is
Wherein,for the policy function value, characterizing the probability of selecting the action parameter a under the state parameter s; q (s, a) is a value of a function, b is a sum index, indicating that all possible traversals are possibleAn action parameter a; τ is a set positive integer constant.
In one possible design, the updating the value function values according to the state set and the reward parameter at the next time includes:
substituting the state set at the next moment and the value function value at the current moment into a preset value function iteration formula to calculate to obtain an updated value function value, wherein the value function iteration formula is that
Wherein Q is new （s t ，a t ) For the updated value function value, Q (s, a) is the value function value at the current time, t represents the current time, t+1 represents the next time, α is the set learning rate, and α is between 0 and 1, γ is the set discount factor, and γ is between 0 and 1,is at the next state parameter s t+1 The maximum value of the function values of the corresponding values of all possible action parameters, r t+1 Characterizing an action parameter a at the current instant t of execution t Then obtaining the state parameter s at the next moment t+1 Is calculated from the reward function.
In one possible design, the excitation function is
r=△s1×ω1+△s2×ω2+△s3×ω3-a1×ω4-|a2|×ω5
Wherein r is a reward parameter, Δs1 is a difference value between a flower opening degree parameter at the next moment and a flower opening degree parameter at the current moment, Δs2 is a difference value between a color vividness parameter at the next moment and a color vividness parameter at the current moment, Δs3 is a difference value between a health state parameter at the next moment and a health state parameter at the current moment, a1 represents a nutrient solution supply adjustment parameter, a2 represents an indoor temperature adjustment parameter, and ω1, ω2, ω3, ω4 and ω5 are set weight coefficients.
In a second aspect, a cut-flower opening cycle optimization system is provided, including an acquisition unit, an extraction unit, a calculation unit, an execution unit, a determination unit, an update unit, and a deployment unit, wherein:
the flower monitoring system comprises an acquisition unit, a control unit and a control unit, wherein the acquisition unit is used for acquiring a flower monitoring data set at the current moment, and the flower monitoring data set comprises a monitoring image of a flower and a plurality of growth environment monitoring parameters; acquiring an action set and an initial value function value table, wherein the action set comprises a plurality of maintenance action parameters, the initial value function value table comprises a plurality of preset value function values, and each value function value is associated with a corresponding state parameter and action parameter; collecting a flower monitoring data set at the next moment;
the extracting unit is used for extracting the characteristics of the monitoring image to obtain a plurality of flower image characteristic parameters, taking each flower image characteristic parameter and each growth environment monitoring parameter as corresponding state parameters, and forming a state set at the current moment by utilizing each state parameter;
the calculation unit is used for calculating the set strategy function based on the initial value function value table and the state set at the current moment to obtain a corresponding strategy function value; calculating the set strategy function based on the updated value function value table and the state set at the next moment to obtain a strategy function value corresponding to the next moment;
the executing unit is used for selecting the action parameters at the next moment from the action set according to the strategy function values, and transmitting the action parameters at the next moment to the cut flower maintenance end so that the cut flower maintenance end executes the action parameters at the next moment to carry out flower maintenance;
the determining unit is used for calculating and determining the rewarding parameter of the next moment by adopting a preset rewarding function based on the state set and the action set of the next moment;
the updating unit is used for updating each value function value according to the state set and the rewarding parameter at the next moment to obtain an updated value function value table;
the deployment unit is used for deploying the strategy function and the final value function value table to the cut flower maintenance end so that the cut flower maintenance end can carry out open period maintenance of the cut flowers by utilizing the strategy function and the final value function value table.
In a third aspect, there is provided a cut-flower opening cycle optimizing apparatus comprising:
a memory for storing instructions;
and a processor for reading the instructions stored in the memory and executing the method according to any one of the above first aspects according to the instructions.
In a fourth aspect, there is provided a computer readable storage medium having instructions stored thereon which, when run on a computer, cause the computer to perform the method of any of the first aspects. Also provided is a computer program product comprising instructions which, when run on a computer, cause the computer to perform the method of any of the first aspects.
The beneficial effects are that: under the iterative coordination of cut flower state monitoring, maintenance action setting and rewarding mechanism, the invention can obtain the optimal management mode of cut flowers through self-adaptive learning by strengthening learning of cut flower maintenance strategies, promote the maintenance effect of cut flowers, optimize and prolong the open period of cut flowers, and enable large-scale personalized fresh keeping of various cut flowers to be possible. The invention can realize automatic maintenance of cut flowers; the state characteristics of flowers can be automatically extracted through image recognition and analysis based on a convolutional neural network; based on the decision-making mode of reinforcement learning, automatic intervention of the whole life cycle of the cut flowers can be realized so as to prolong the flowering period of the cut flowers, and meanwhile, the cost for maintaining the cutting flower opening period is considered.
Drawings
In order to more clearly illustrate the embodiments of the invention or the technical solutions in the prior art, the drawings that are required in the embodiments or the description of the prior art will be briefly described, it being obvious that the drawings in the following description are only some embodiments of the invention, and that other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.
FIG. 1 is a schematic diagram showing the steps of the method of example 1 of the present invention;
FIG. 2 is a schematic diagram showing the construction of a system in embodiment 2 of the present invention;
fig. 3 is a schematic view showing the constitution of the apparatus in embodiment 3 of the present invention.
Detailed Description
It should be noted that the description of these examples is for aiding in understanding the present invention, but is not intended to limit the present invention. Specific structural and functional details disclosed herein are merely representative of example embodiments of the invention. This invention may, however, be embodied in many alternate forms and should not be construed as limited to the embodiments set forth herein.
It will be appreciated that the term "coupled" is to be interpreted broadly, and may be a fixed connection, a removable connection, or an integral connection, for example, unless explicitly stated and limited otherwise; can be mechanically or electrically connected; can be directly connected or indirectly connected through an intermediate medium, and can be communication between two elements. The specific meaning of the terms in the embodiments can be understood by those of ordinary skill in the art according to the specific circumstances.
In the following description, specific details are provided to provide a thorough understanding of example embodiments. However, it will be understood by those of ordinary skill in the art that the example embodiments may be practiced without these specific details. For example, a system may be shown in block diagrams in order to avoid obscuring the examples with unnecessary detail. In other embodiments, well-known processes, structures, and techniques may be shown without unnecessary detail in order to avoid obscuring the embodiments.
Example 1:
the embodiment provides a cut flower opening period optimization method, which can be applied to a corresponding cut flower maintenance system, as shown in fig. 1, and comprises the following steps:
s1, acquiring a flower monitoring data set at the current moment, wherein the flower monitoring data set comprises a flower monitoring image and a plurality of growth environment monitoring parameters.
In specific implementation, the monitoring images of flowers can be collected regularly by using the cameras, the growth environment monitoring parameters of the flowers are collected regularly by using the corresponding temperature and humidity sensor and the corresponding illumination sensor, and the cut flower maintenance end forms a flower monitoring data set by using the monitoring images and the growth environment monitoring parameters. The system can acquire the flower monitoring data set at the current moment from the cut flower maintenance end.
S2, extracting features of the monitoring images to obtain a plurality of flower image feature parameters, taking the flower image feature parameters and the growth environment monitoring parameters as corresponding state parameters, and forming a state set at the current moment by utilizing the state parameters.
In specific implementation, the system can input the monitoring image into a convolutional neural network model trained by a training set to extract image features to obtain a plurality of flower image feature parameters, wherein the training set comprises a plurality of flower image samples marked with corresponding image feature quantization indexes. The flower image characteristic parameters comprise flower opening degree parameters (which can be quantized to the degree of petal unfolding and are divided into 0-9 ten grade parameters), color vividness parameters (which can be quantized to the color saturation and brightness of the flower and are divided into 0-9 ten grade parameters) and health state parameters (which can be quantized to the degree of leaf withering of the flower and are divided into 0-9 ten grade parameters), and the growth environment monitoring parameters comprise indoor temperature parameters, relative humidity parameters and illumination intensity parameters. The system takes the characteristic parameters of each flower image and the monitoring parameters of each growth environment as corresponding state parameters, and utilizes each state parameter to form a state set at the current moment.
Illustratively, the training process of the convolutional neural network model includes:
a number of flower images are acquired. Various flowers can be shot by using the high-resolution camera, so that the image quality is clear and rich flower characteristic information is contained.
And (3) carrying out corresponding image characteristic quantization index labeling on each flower image, wherein the image characteristic quantization index labeling comprises flower type labeling, flower openness quantization index (petal unfolding degree is divided into 0-9 grade parameters), color vividness quantization index (color saturation and brightness of the flower are divided into 0-9 grade parameters) labeling and health state quantization index (flower leaf withering degree is divided into 0-9 grade parameters) labeling, and the labeled flower images are utilized to form an initial image set.
Performing image preprocessing and image enhancement processing on each flower image in the initial image set to obtain a training set; image preprocessing includes image size adjustment, image denoising, image normalization processing, and the like, and image enhancement processing includes rotation, clipping, flipping, affine transformation, and the like, to expand the data set scale and enhance the robustness of the trained model.
Training a convolutional neural network model constructed in advance by using a training set to obtain a trained convolutional neural network model. The cross entropy loss function is used for flower classification, assuming N samples, each sample having K number The cross entropy is
For measuring the difference between predicted and actual values, where y i，k True tags representing the kth class of the ith sample (1 if belonging to that class, or 0 otherwise), p i，k Is the predicted probability that the model belongs to the kth class for the ith sample.
The root mean square error of the predicted and actual values is used as a loss function for the optimal environmental parameters. And employs a random gradient descent (SGD) or Adam optimizer to minimize the loss function. The performance of the model is monitored by verifying the accuracy (for flower variety and status classification) on the set to obtain an optimal flower variety and status classification predictive model.
The convolutional neural network (Convolutional Neural Network, CNN) is a deep learning model, is excellent in image recognition task, and realizes the following technology aiming at flower varieties and states:
and inputting a data representation. The picture data of flowers are expressed in the form of a matrix, and three-dimensional parameters of the matrix respectively comprise width, height and color channel number.
And processing input data. The input data undergoes a series of transformation operations:
first is convolution
Then activate
Followed by pooling
Then a full connection layer calculation (where X is the input feature, W is the weight parameter, B is the bias.)
Finally, the Softmax function gives the probability of a particular class (here represented by j)
For example, when the input image matrix is
Within the convolutional layer, the convolutional kernel is a small matrix that is a learnable parameter of the convolutional layer. Typically, the size of the convolution kernel is much smaller than the size of the input image. For example, the convolution kernel is
The convolution kernel performs a sliding window operation on the input image. The sliding window is moved in fixed steps over the input image and performs a dot product operation with the local area at each position and sums the results to obtain a pixel value of the output feature map. The above steps are repeated until the entire input image is traversed. This will generate an output signature whose size may be different from the input image, depending on the size and stride of the convolution kernel. Such as convolution operations as
Next, an activation operation is performed using the ReLU function
Then carrying out maximum pooling operation, dividing the matrix into blocks taking the 2×2 matrix as a unit, taking the maximum value of the matrix in each block, and constructing a new matrix with smaller scale
Then, a full connection layer is constructed, data calculation is carried out, and the weight matrix of the connection layer is as follows
Further set a bias term
The available full join operation is
The result of the full join operation
Input Softmax function
During practical training, a more complex CNN network can be constructed by superposing a plurality of convolution layers and full connection layers, and then a convolution kernel K matrix, a full connection layer parameter W matrix and a bias term B are adjusted by taking the minimum cross entropy as a target through a given flower classification data set.
S3, acquiring an action set and an initial value function value table, wherein the action set comprises a plurality of maintenance action parameters, the initial value function value table comprises a plurality of preset value function values, and each value function value is associated with a corresponding state parameter and action parameter.
In specific implementation, the system further obtains an action set and an initial value function value table, wherein the action set comprises a plurality of maintenance action parameters, including a moisture supply adjustment parameter (which can be quantized into milliliters of added water), a nutrient solution supply adjustment parameter (which can be quantized into milliliters of added nutrient solution), an illumination intensity adjustment parameter (which can be quantized into a light-on time period), an indoor temperature adjustment parameter (which can be quantized into a heating or refrigerating time period) and an indoor humidity adjustment parameter (which can be quantized into a humidifier-on time period). The initial value function value table comprises a plurality of preset value function values, each value function value is associated with a corresponding state parameter and action parameter, the initial value function value table is gradually updated according to subsequent reinforcement learning experience, and the value function value in the initial value function value table can be randomly smaller.
S4, calculating the set strategy function based on the initial value function value table and the state set at the current moment to obtain the corresponding strategy function value.
In specific implementation, the system can calculate the set strategy function based on the initial value function value table and the state set at the current moment to obtain the corresponding strategy function value. The policy function is
Wherein,for the policy function value, characterizing the probability of selecting the action parameter a under the state parameter s; q (s, a) is a value function value, b is a sum index, and represents traversing all possible action parameters a; τ is a set positive integer constant, which can be set as a positive number of temperature parameters, which controls the degree of exploration of the strategy, the higher the temperature, the more equally the strategy tends to explore all actions, and the lower the temperature, the more the strategy tends to select the action with the highest action value.
S5, selecting action parameters at the next moment from the action set according to the strategy function values, and transmitting the action parameters at the next moment to the cut flower maintenance end, so that the cut flower maintenance end executes the action parameters at the next moment to carry out flower maintenance.
In particular, the system selects the action parameters of the next moment in the action set according to the calculated policy function values and the probability, for example, if three actions A1, A2 and A3 are selected according to the probabilities of pi 1, pi 2 and pi 3, then a number between 0 and 1 is randomly generated, then if the number falls between [0 and pi 1], the action A1 is selected, if the number falls between (pi 1 and pi 2], the action A2 is selected, if the number falls between (pi 1 and pi 2 and pi 1), the action parameter of the next moment is selected to be transmitted to the cut flower maintenance end by the system, and the like.
S6, collecting a flower monitoring data set at the next moment, and executing a step S2 on the flower monitoring data set at the next moment to obtain a state set at the next moment.
In specific implementation, the system needs to observe a new state of the flowers after corresponding action parameters are executed, a flower monitoring data set at the next moment can be collected from the cut flower maintenance end, and the flower monitoring data set at the next moment is processed in the step S2 to obtain a state set at the next moment.
S7, calculating and determining the rewarding parameter of the next moment by adopting a preset rewarding function based on the state set and the action set of the next moment.
In particular, the system calculates and determines the prize parameters of the next moment by using a preset prize function according to the state set and the action set of the next moment, and the prize parameter is exemplified by the incentive function
r=△s1×ω1+△s2×ω2+△s3×ω3-a1×ω4-|a2|×ω5
Wherein r is a reward parameter, Δs1 is a difference value between a flower opening degree parameter at the next moment and a flower opening degree parameter at the current moment, Δs2 is a difference value between a color vividness parameter at the next moment and a color vividness parameter at the current moment, Δs3 is a difference value between a health state parameter at the next moment and a health state parameter at the current moment, a1 represents a nutrient solution supply adjustment parameter, a2 represents an indoor temperature adjustment parameter, ω1, ω2, ω3, ω4 and ω5 are set weight coefficients, the relative importance of different factors in the reward function is determined, and the values of the weight coefficients need to be adjusted according to actual conditions so as to ensure that an algorithm can correctly reflect the preference and actual operation cost of a user, for example, the consumption of the nutrient solution is required to be reduced, and the value of ω 4 is increased. The nutrient solution supply adjustment parameters (which can be quantized into milliliters of added nutrient solution) and the illumination intensity adjustment parameters (which can be quantized into the duration of turning on the lamp) are selected according to the action set in the excitation function, because the cost of maintaining the cutting flower on-period is needed to be considered due to the fact that the cost is the most costly, other actions such as adding water can indirectly influence the opening state change of flowers, and therefore the rewarding function is indirectly influenced.
S8, updating each value function value according to the state set and the rewarding parameter at the next moment to obtain an updated value function value table.
In specific implementation, the system updates each value function value according to the state set and the rewarding parameter at the next moment, and specifically comprises the following steps:
substituting the state set at the next moment and the value function value at the current moment into a preset value function iteration formula to calculate to obtain an updated value function value, wherein the value function iteration formula is that
Wherein Q is new （s t ，a t ) To update the value function value, Q (s, a) is the same asThe value of the function value of the previous moment, t represents the current moment, t+1 represents the next moment, alpha is the set learning rate, alpha is between 0 and 1, gamma is the set discount factor, gamma is between 0 and 1,is at the next state parameter s t+1 The maximum value of the function values of the corresponding values of all possible action parameters, r t+1 Characterizing an action parameter a at the current instant t of execution t Then obtaining the state parameter s at the next moment t+1 Is calculated from the reward function.
S9, calculating the set strategy function based on the updated value function value table and the state set at the next moment to obtain the strategy function value corresponding to the next moment.
When the method is implemented, the system calculates the strategy function again based on the updated value function value table and the state set of the next moment to obtain the strategy function value corresponding to the next moment.
S10, iteratively executing the steps S5-S9 based on the strategy function value corresponding to the next moment until the set iteration condition is met, and obtaining a final value function value table.
In specific implementation, the system iteratively executes steps S5-S9 based on the strategy function value corresponding to the next moment until the set iteration condition is met, so as to obtain a final value function value table, wherein the iteration condition can be set as that the flower maintenance reaches the preset flowering period length or enough iteration times are passed.
S11, deploying the strategy function and the final value function value table to the cut flower maintenance end so that the cut flower maintenance end performs cut flower open period maintenance by using the strategy function and the final value function value table.
When the method is implemented, after strategy reinforcement learning meets the set iteration conditions, a good strategy is found, a strategy test can be performed to see whether the method can effectively prolong the flowering period of the cut flowers under the actual conditions, if the test result is satisfactory, the strategy, namely a strategy function and a final value function value, can be deployed to an automatic cut flower maintenance end, so that the cut flower maintenance end uses the strategy function and the final value function value table to perform open period maintenance of corresponding cut flowers.
In the method, the optimal management mode of the cut flowers can be obtained by self-adaptive learning, the maintenance effect of the cut flowers is improved, the opening period of the cut flowers is prolonged, and large-scale personalized fresh-keeping of various cut flowers is possible. The method can realize automatic maintenance of cut flowers; the state characteristics of flowers can be automatically extracted through image recognition and analysis based on a convolutional neural network; based on the decision-making mode of reinforcement learning, the automatic intervention of the whole life cycle of the cut flowers can be realized, so that the flowering period of the cut flowers can be prolonged.
Example 2:
the embodiment provides a cut flower opening period optimizing system, as shown in fig. 2, including an obtaining unit, an extracting unit, a calculating unit, an executing unit, a determining unit, an updating unit and a deploying unit, wherein:
the flower monitoring system comprises an acquisition unit, a control unit and a control unit, wherein the acquisition unit is used for acquiring a flower monitoring data set at the current moment, and the flower monitoring data set comprises a monitoring image of a flower and a plurality of growth environment monitoring parameters; acquiring an action set and an initial value function value table, wherein the action set comprises a plurality of maintenance action parameters, the initial value function value table comprises a plurality of preset value function values, and each value function value is associated with a corresponding state parameter and action parameter; collecting a flower monitoring data set at the next moment;
the extracting unit is used for extracting the characteristics of the monitoring image to obtain a plurality of flower image characteristic parameters, taking each flower image characteristic parameter and each growth environment monitoring parameter as corresponding state parameters, and forming a state set at the current moment by utilizing each state parameter;
the calculation unit is used for calculating the set strategy function based on the initial value function value table and the state set at the current moment to obtain a corresponding strategy function value; calculating the set strategy function based on the updated value function value table and the state set at the next moment to obtain a strategy function value corresponding to the next moment;
the executing unit is used for selecting the action parameters at the next moment from the action set according to the strategy function values, and transmitting the action parameters at the next moment to the cut flower maintenance end so that the cut flower maintenance end executes the action parameters at the next moment to carry out flower maintenance;
the determining unit is used for calculating and determining the rewarding parameter of the next moment by adopting a preset rewarding function based on the state set and the action set of the next moment;
the updating unit is used for updating each value function value according to the state set and the rewarding parameter at the next moment to obtain an updated value function value table;
the deployment unit is used for deploying the strategy function and the final value function value table to the cut flower maintenance end so that the cut flower maintenance end can carry out open period maintenance of the cut flowers by utilizing the strategy function and the final value function value table.
Example 3:
the embodiment provides a cut-flower opening period optimizing device, as shown in fig. 3, including, at a hardware level:
the data interface is used for establishing data butt joint between the processor and the cut flower maintenance end;
a memory for storing instructions;
and the processor is used for reading the instructions stored in the memory and executing the cutting flower opening period optimizing method in the embodiment 1 according to the instructions.
Optionally, the device further comprises an internal bus. The processor and memory and data interfaces may be interconnected by an internal bus, which may be an ISA (Industry Standard Architecture ) bus, a PCI (Peripheral Component Interconnect, peripheral component interconnect standard) bus, or an EISA (Extended Industry Standard Architecture ) bus, among others. The buses may be classified as address buses, data buses, control buses, etc.
The Memory may include, but is not limited to, random access Memory (Random Access Memory, RAM), read Only Memory (ROM), flash Memory (Flash Memory), first-in first-out Memory (First Input First Output, FIFO), and/or first-in last-out Memory (First In Last Out, FILO), etc. The processor may be a general-purpose processor including a central processing unit (Central Processing Unit, CPU), a network processor (Network Processor, NP), etc.; but also digital signal processors (Digital Signal Processor, DSP), application specific integrated circuits (Application Specific Integrated Circuit, ASIC), field programmable gate arrays (Field-Programmable Gate Array, FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components.
Example 4:
the present embodiment provides a computer-readable storage medium having instructions stored thereon that, when executed on a computer, cause the computer to perform the cut-flower opening cycle optimization method of embodiment 1. The computer readable storage medium refers to a carrier for storing data, and may include, but is not limited to, a floppy disk, an optical disk, a hard disk, a flash Memory, and/or a Memory Stick (Memory Stick), etc., where the computer may be a general purpose computer, a special purpose computer, a computer network, or other programmable system.
The present embodiment also provides a computer program product comprising instructions which, when run on a computer, cause the computer to perform the cut-flower opening cycle optimization method of embodiment 1. Wherein the computer may be a general purpose computer, a special purpose computer, a computer network, or other programmable system.
Finally, it should be noted that: the foregoing description is only of the preferred embodiments of the invention and is not intended to limit the scope of the invention. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.
Claims (10)
1. A cut-flower opening cycle optimization method, comprising:
a. acquiring a flower monitoring data set at the current moment, wherein the flower monitoring data set comprises a flower monitoring image and a plurality of growth environment monitoring parameters;
b. extracting features of the monitoring images to obtain a plurality of flower image feature parameters, taking each flower image feature parameter and each growth environment monitoring parameter as corresponding state parameters, and forming a state set at the current moment by utilizing each state parameter;
c. acquiring an action set and an initial value function value table, wherein the action set comprises a plurality of maintenance action parameters, the initial value function value table comprises a plurality of preset value function values, and each value function value is associated with a corresponding state parameter and action parameter;
d. calculating a set strategy function based on the initial value function value table and the state set at the current moment to obtain a corresponding strategy function value;
e. selecting action parameters of the next moment from the action set according to the strategy function values, and transmitting the action parameters of the next moment to the cut flower maintenance end, so that the cut flower maintenance end executes the action parameters of the next moment to carry out flower maintenance;
f. collecting a flower monitoring data set at the next moment, and executing the step b on the flower monitoring data set at the next moment to obtain a state set at the next moment;
g. calculating and determining a reward parameter of the next moment by adopting a preset reward function based on the state set and the action set of the next moment;
h. updating each value function value according to the state set and the rewarding parameter at the next moment to obtain an updated value function value table;
i. calculating a set strategy function based on the updated value function value table and a state set at the next moment to obtain a strategy function value corresponding to the next moment;
j. iteratively executing the steps e-i based on the strategy function value corresponding to the next moment until the set iteration condition is met, so as to obtain a final value function value table;
k. and deploying the strategy function and the final value function value table to the cut flower maintenance end so that the cut flower maintenance end performs cut flower open period maintenance by using the strategy function and the final value function value table.
2. The method for optimizing a cut-flower opening cycle according to claim 1, wherein the feature extraction of the monitored image to obtain a plurality of flower image feature parameters comprises:
and inputting the monitoring image into a convolutional neural network model trained by a training set to extract image features to obtain a plurality of flower image feature parameters, wherein the training set comprises a plurality of flower image samples marked with corresponding image feature quantization indexes.
3. A cut-flower opening cycle optimization method according to claim 2, characterized in that before feature extraction of the monitored image, the method further comprises:
collecting a plurality of flower images;
corresponding image characteristic quantization index labeling is carried out on each flower image, wherein the image characteristic quantization index labeling comprises flower openness quantization index labeling, color vividness quantization index labeling and health state quantization index labeling, and each flower image after labeling is utilized to form an initial image set;
performing image preprocessing and image enhancement processing on each flower image in the initial image set to obtain a training set;
training a convolutional neural network model constructed in advance by using a training set to obtain a trained convolutional neural network model.
4. The method of optimizing a cut-flower opening cycle according to claim 1, wherein the plurality of growth environment monitoring parameters includes an indoor temperature parameter, a relative humidity parameter, and an illumination intensity parameter; the flower image characteristic parameters comprise flower opening degree parameters, color vividness parameters and health state parameters; the plurality of maintenance action parameters comprise a water supply adjustment parameter, a nutrient solution supply adjustment parameter, an illumination intensity adjustment parameter, an indoor temperature adjustment parameter and an indoor humidity adjustment parameter.
5. The method for optimizing cut open period as claimed in claim 4, wherein said strategy function is
Wherein,for the policy function value, characterizing the probability of selecting the action parameter a under the state parameter s; q (s, a) is a value function value, b is a sum index, and represents traversing all possible action parameters a; τ is a set positive integer constant.
6. The method of optimizing a cut-flower opening cycle according to claim 4, wherein updating the value function values based on the state set and the bonus parameter at the next time includes:
substituting the state set at the next moment and the value function value at the current moment into a preset value function iteration formula to calculate to obtain an updated value function value, wherein the value function iteration formula is that
Wherein Q is new （s t ，a t ) For the updated value function value, Q (s, a) is the value function value at the current time, t represents the current time, t+1 represents the next time, α is the set learning rate, and α is between 0 and 1, γ is the set discount factor, and γ is between 0 and 1,is at the next state parameter s t+1 The maximum value of the function values of the corresponding values of all possible action parameters, r t+1 Characterizing an action parameter a at the current instant t of execution t Then obtaining the state parameter s at the next moment t+1 Are awarded parameters of (a)Calculated from the reward function.
7. The method of optimizing cut open period as claimed in claim 6, wherein said excitation function is
r=△s1×ω1+△s2×ω2+△s3×ω3-a1×ω4-|a2|×ω5
Wherein r is a reward parameter, Δs1 is a difference value between a flower opening degree parameter at the next moment and a flower opening degree parameter at the current moment, Δs2 is a difference value between a color vividness parameter at the next moment and a color vividness parameter at the current moment, Δs3 is a difference value between a health state parameter at the next moment and a health state parameter at the current moment, a1 represents a nutrient solution supply adjustment parameter, a2 represents an indoor temperature adjustment parameter, and ω1, ω2, ω3, ω4 and ω5 are set weight coefficients.
8. The system for optimizing the cutting and opening period is characterized by comprising an acquisition unit, an extraction unit, a calculation unit, an execution unit, a determination unit, an updating unit and a deployment unit, wherein:
the flower monitoring system comprises an acquisition unit, a control unit and a control unit, wherein the acquisition unit is used for acquiring a flower monitoring data set at the current moment, and the flower monitoring data set comprises a monitoring image of a flower and a plurality of growth environment monitoring parameters; acquiring an action set and an initial value function value table, wherein the action set comprises a plurality of maintenance action parameters, the initial value function value table comprises a plurality of preset value function values, and each value function value is associated with a corresponding state parameter and action parameter; collecting a flower monitoring data set at the next moment;
the extracting unit is used for extracting the characteristics of the monitoring image to obtain a plurality of flower image characteristic parameters, taking each flower image characteristic parameter and each growth environment monitoring parameter as corresponding state parameters, and forming a state set at the current moment by utilizing each state parameter;
the calculation unit is used for calculating the set strategy function based on the initial value function value table and the state set at the current moment to obtain a corresponding strategy function value; calculating the set strategy function based on the updated value function value table and the state set at the next moment to obtain a strategy function value corresponding to the next moment;
the executing unit is used for selecting the action parameters at the next moment from the action set according to the strategy function values, and transmitting the action parameters at the next moment to the cut flower maintenance end so that the cut flower maintenance end executes the action parameters at the next moment to carry out flower maintenance;
the determining unit is used for calculating and determining the rewarding parameter of the next moment by adopting a preset rewarding function based on the state set and the action set of the next moment;
the updating unit is used for updating each value function value according to the state set and the rewarding parameter at the next moment to obtain an updated value function value table;
the deployment unit is used for deploying the strategy function and the final value function value table to the cut flower maintenance end so that the cut flower maintenance end can carry out open period maintenance of the cut flowers by utilizing the strategy function and the final value function value table.
9. A cut-flower opening cycle optimizing apparatus, characterized by comprising:
a memory for storing instructions;
a processor for reading the instructions stored in the memory and executing the cut open period optimization method according to any one of claims 1-7 according to the instructions.
10. A computer readable storage medium having instructions stored thereon which, when executed on a computer, cause the computer to perform the cut-flower opening cycle optimization method of any one of claims 1-7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410002772.2A CN117789095B (en) | 2024-01-02 | 2024-01-02 | Cut flower opening period optimization method, system, equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410002772.2A CN117789095B (en) | 2024-01-02 | 2024-01-02 | Cut flower opening period optimization method, system, equipment and storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN117789095A true CN117789095A (en) | 2024-03-29 |
| CN117789095B CN117789095B (en) | 2024-05-14 |
Family
ID=90390869
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202410002772.2A Active CN117789095B (en) | 2024-01-02 | 2024-01-02 | Cut flower opening period optimization method, system, equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117789095B (en) |
Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20170094920A1 (en) * | 2015-10-02 | 2017-04-06 | Craig Ellins | Integrated incubation, cultivation and curing system and controls for optimizing and enhancing plant growth, development and performance of plant-based medical therapies |
| CN108629690A (en) * | 2018-04-28 | 2018-10-09 | 福州大学 | Futures based on deeply study quantify transaction system |
| CN110199844A (en) * | 2019-05-07 | 2019-09-06 | 青岛科技大学 | Potted plant intelligence curing system and its data processing method based on machine vision |
| CN110598841A (en) * | 2018-06-13 | 2019-12-20 | 南京大学 | Flower disease analysis method based on multi-input convolutional neural network |
| KR102303169B1 (en) * | 2021-02-24 | 2021-09-17 | 주식회사 아이티로 | Artificial intelligence concrete wet curing management system |
| CN114282738A (en) * | 2020-09-27 | 2022-04-05 | 南通秉匠信息科技有限公司 | Garden vegetation informatization management method and device based on three-dimensional visualization, terminal and medium |
| WO2022241808A1 (en) * | 2021-05-19 | 2022-11-24 | 广州中国科学院先进技术研究所 | Multi-robot trajectory planning method |
| US20230082326A1 (en) * | 2020-02-07 | 2023-03-16 | Deepmind Technologies Limited | Training multi-objective neural network reinforcement learning systems |
| CN116882297A (en) * | 2023-07-25 | 2023-10-13 | 国网浙江省电力有限公司杭州供电公司 | Shaft hole assembly strategy model construction method, device, equipment and medium |
| WO2023216861A1 (en) * | 2022-05-11 | 2023-11-16 | 杭州睿胜软件有限公司 | Personalized plant care method and system, and readable storage medium |
-
2024
- 2024-01-02 CN CN202410002772.2A patent/CN117789095B/en active Active
Patent Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20170094920A1 (en) * | 2015-10-02 | 2017-04-06 | Craig Ellins | Integrated incubation, cultivation and curing system and controls for optimizing and enhancing plant growth, development and performance of plant-based medical therapies |
| CN108629690A (en) * | 2018-04-28 | 2018-10-09 | 福州大学 | Futures based on deeply study quantify transaction system |
| CN110598841A (en) * | 2018-06-13 | 2019-12-20 | 南京大学 | Flower disease analysis method based on multi-input convolutional neural network |
| CN110199844A (en) * | 2019-05-07 | 2019-09-06 | 青岛科技大学 | Potted plant intelligence curing system and its data processing method based on machine vision |
| US20230082326A1 (en) * | 2020-02-07 | 2023-03-16 | Deepmind Technologies Limited | Training multi-objective neural network reinforcement learning systems |
| CN114282738A (en) * | 2020-09-27 | 2022-04-05 | 南通秉匠信息科技有限公司 | Garden vegetation informatization management method and device based on three-dimensional visualization, terminal and medium |
| KR102303169B1 (en) * | 2021-02-24 | 2021-09-17 | 주식회사 아이티로 | Artificial intelligence concrete wet curing management system |
| WO2022241808A1 (en) * | 2021-05-19 | 2022-11-24 | 广州中国科学院先进技术研究所 | Multi-robot trajectory planning method |
| WO2023216861A1 (en) * | 2022-05-11 | 2023-11-16 | 杭州睿胜软件有限公司 | Personalized plant care method and system, and readable storage medium |
| CN116882297A (en) * | 2023-07-25 | 2023-10-13 | 国网浙江省电力有限公司杭州供电公司 | Shaft hole assembly strategy model construction method, device, equipment and medium |
Non-Patent Citations (2)
| Title |
|---|
| 杨静亚;李景霞;王振宇;程海;: "基于卷积神经网络的花朵品种的识别", 黑龙江大学工程学报, no. 04, 25 December 2019 (2019-12-25) * |
| 黄宏梅;陆卫忠;杨茹;曹燕;: "深度学习在花卉养护技术中的应用", 计算机系统应用, no. 03, 15 March 2020 (2020-03-15) * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN117789095B (en) | 2024-05-14 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| WO2022253057A1 (en) | Intelligent moisture precise irrigation control system and method for fruit and vegetable cultivation in solar greenhouse | |
| CN106971152B (en) | Method for detecting bird nest in power transmission line based on aerial images | |
| CN112418498B (en) | Temperature prediction method and system for intelligent greenhouse | |
| CN110781595B (en) | Method, device, terminal and medium for predicting energy use efficiency (PUE) | |
| CN116058195B (en) | Illumination regulation and control method, system and device for leaf vegetable growth environment | |
| CN111639750A (en) | Control method and device of intelligent flowerpot, intelligent flowerpot and storage medium | |
| CN113034301A (en) | Crop growth management system and method | |
| CN112272074B (en) | Information transmission rate control method and system based on neural network | |
| CN110648020A (en) | Greenhouse crop water demand prediction method and device | |
| CN110866872A (en) | Pavement crack image preprocessing intelligent selection method and device and electronic equipment | |
| WO2023024615A1 (en) | System and method for performing precise irrigation on the basis of plant wilting degree | |
| CN114879786A (en) | Method, system, device and medium for acquiring edible fungus decision scheme | |
| CN111027436A (en) | Northeast black fungus disease and pest image recognition system based on deep learning | |
| CN117789095B (en) | Cut flower opening period optimization method, system, equipment and storage medium | |
| CN114115394A (en) | Intelligent tobacco leaf baking control system and method | |
| CN116863403B (en) | Crop big data environment monitoring method and device and electronic equipment | |
| CN117252344A (en) | Suitability evaluation method and device for cultivation facility, electronic equipment and storage medium | |
| CN117253192A (en) | Intelligent system and method for silkworm breeding | |
| CN116341391B (en) | Precipitation prediction method based on STPM-XGBoost model | |
| CN112464172A (en) | Growth parameter active and passive remote sensing inversion method and device | |
| CN116630080A (en) | Method and system for determining capacity of aquatic product intensive culture feed based on image recognition | |
| CN108845075B (en) | Compost maturity real-time prediction method based on deep learning network | |
| CN116597309A (en) | Method for identifying and selecting plant resistant varieties by integrating physiological indexes of germination period growth and seedling period | |
| CN112443943A (en) | Model training method based on small amount of labeled data, control system and air conditioner | |
| CN115510991A (en) | Sunlight greenhouse aphid early warning method based on 1DCNN-LSTM multi-source information fusion model |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |