Deinsectization device and method

Abstract

The embodiment of the invention provides a deinsectization device and method. The deinsectization device and method have the following beneficial effects that the device comprises a base, a supporting assembly, a shell, a collecting plate, an electronic control module and a sensor module; the base is movably connected with the shell through the supporting assembly; the shell is provided with an insect trapping assembly and an insect killing assembly, and the shell comprises a concave cavity which forms a high-temperature cavity during working; the collecting plate is arranged on the supporting assembly; the sensor module comprises a temperature and humidity sensor, an illumination sensor, a wind speed sensor and a camera which are all in communication connection with the electronic control module; the electronic control module adjusts control parameters of the insect trapping assembly and the insect killing assembly according to the collected temperature, humidity, light intensity, wind speed and images. The environment sensing technology, the multi-light-source combination and the self-adaptive adjustment algorithm are combined, the excellent pest prevention and control effect can be kept in various complex environments, energy consumption can be remarkably reduced, the adverse effect on the ecological environment is reduced, the pest killing efficiency is effectively improved, and high environment adaptability is achieved.

Description

Technical Field

[0001] The present invention relates to the field of agricultural pest control, and in particular to an insect extermination device and method. Background Art

[0002] Agricultural pest control is an important task in agricultural production and is related to the yield and quality of crops. Traditional pest control methods mainly rely on chemical pesticides. Although they can effectively control pests in the short term, long-term use will increase pest resistance. In recent years, with the enhancement of environmental awareness and the advancement of science and technology, finding efficient and environmentally friendly pest control methods has become a research hotspot. Among them, high-temperature killing technology based on light sources has gradually attracted attention due to its advantages such as no pollution and low energy consumption. Existing technologies are mostly based on a fixed combination of light sources and heat sources, but in actual applications, due to changes in environmental conditions, the effect of the equipment is often not optimal. Different environmental factors, such as temperature, humidity, light intensity and wind speed, will affect the efficiency and energy consumption of the insecticide.

[0003] From the above description, it can be seen that it is difficult to achieve flexible adaptation to the environment in current agricultural pest control. How to improve pest control efficiency and environmental adaptability is a technical problem that technical personnel in this field urgently need to solve. Summary of the invention

[0004] In order to overcome the shortcomings of low efficiency and poor environmental adaptability of existing pest control methods, the present invention proposes an pest control device and method.

[0005] In order to achieve the above-mentioned object, according to a first aspect of the present invention, an embodiment of the present invention provides an insect-killing device, which includes a base, a support assembly, a housing, a collection plate, an electronic control module and a sensor module:

[0006] The base is movably connected to the shell through a support assembly;

[0007] The housing is provided with an insect attracting component and an insect killing component, and the housing includes a concave cavity for forming a high-temperature cavity during operation;

[0008] The collecting plate is arranged on the supporting assembly and is used to collect insects;

[0009] The sensor module includes a temperature and humidity sensor, a light sensor, a wind speed sensor and a camera; the temperature and humidity sensor, the light sensor, the wind speed sensor and the camera are communicatively connected with the electronic control module; the temperature and humidity sensor is used to collect the temperature and humidity of the environment, the light sensor is used to collect the light intensity of the environment, the wind speed sensor is used to collect the wind speed of the environment, and the camera is used to collect the image of the collection plate; the electronic control module is communicatively connected with the insect attracting component and the insect killing component, and the control parameters of the insect attracting component and the insect killing component are calculated and adjusted through a reinforcement learning algorithm according to the collected temperature, humidity, light intensity, wind speed and image.

[0010] Optionally, the temperature and humidity sensor and the light sensor are arranged on the top surface of the base, the wind speed sensor is arranged on the shell, and the camera is arranged above the support assembly and the collection plate to collect images on the collection plate.

[0011] Optionally, the insect attracting component includes a high-pressure mercury lamp, which is arranged at the center of the inner wall of the concave cavity; the insect killing component includes an iodine tungsten lamp, which is arranged at the edge of the inner wall of the concave cavity.

[0012] Optionally, the support assembly includes a support rod and a bracket, one end of the support rod is arranged on the top surface of the base, and the other end is connected to the bracket; the bracket is movably connected to the shell through a rotating shaft so that the shell can rotate by an angle:

[0013] According to a second aspect of the present invention, an embodiment of the present invention further provides an insect killing method based on the insect killing device in the above implementation, the method comprising:

[0014] The temperature, humidity, light intensity and wind speed are obtained through the temperature and humidity sensor, light sensor and wind speed sensor, and form the environmental state vector;

[0015] Input the environmental state vector into the strategy network to calculate the control parameters, wherein the control parameters include the strength of the insect attracting component and the insect killing component, the temperature of the insect killing component, and the band of the insect attracting component;

[0016] The reward value is calculated by weighted calculation based on the capture rate, energy consumption and ecological impact index, wherein the capture rate is the ratio of the number of target organisms captured per unit time to the total capture obtained by analyzing the images collected by the camera, the energy consumption is the energy consumption of the insecticide, and the ecological impact index is the ratio of the number of non-target organisms captured to the total capture;

[0017] Calculate the advantage function value according to the calculated reward value and the value estimate corresponding to the environment state vector at different times, wherein the value estimate is a scalar value calculated by the evaluation network through a multi-layer neural network according to the input environment state vector at different times;

[0018] The evaluation network inputs the calculated advantage function value into the strategy network, and the strategy network adjusts the control parameter according to the advantage function value.

[0019] Optionally, the environment state vector is input into a strategy network to calculate a control parameter, and the formula is as follows:

[0020] a(t)=σ(MLP(s(t)))=[P light (t),T heat (t),B(t)]

[0021] Among them, a(t) represents the control parameter, σ represents the Sigmoid activation function, MLP represents full connection, s(t) represents the environment state vector at time t, P light (t) represents the calculated intensity at time t, T heat (t) represents the temperature at time t, B(t) represents the band at time t, and the strategy network includes an Actor network.

[0022] Optionally, the reward value is weighted and calculated according to the capture rate, energy consumption and ecological impact index, and the formula is as follows:

[0023] B(t) = α·capture rate(t)-β·energy consumption(t)-γ·ecological impact index(t)

[0024] Among them, R(t) represents the reward value at time t, capture rate(t) represents the capture rate at time t, energy consumption(t) represents the energy consumption at time t, ecological impact index(t) represents the ecological impact index at time t, α represents the capture rate weight, β represents the energy consumption weight, and γ represents the ecological impact weight.

[0025] Optionally, the capture rate weight, the energy consumption weight and the ecological impact weight are as follows:

[0026] α+β+γ=1

[0027] 0<α, β, γ<1

[0028] α∈[0.4,0.6], β∈[0.3,0.5], γ∈[0.1,0.3].

[0029] Optionally, the advantage function value is calculated based on the calculated reward value and the value estimate corresponding to the environment state vector at different times, and the formula is as follows:

[0030] A(t)=R(t)+δ·V φ (s(t+1))-V φ (s(t))

[0031] Among them, A(t) represents the advantage function value at time t, R(t) represents the reward value at time t, and V φ (s(t+1)) represents the value estimate at time t+1, s(t+1) represents the environment state vector at time t+1, V φ (s(t)) represents the value estimate at time t, s(t) represents the environment state vector at time t, δ is the discount factor, and 0≤δ<1, and the evaluation network includes a Critic network.

[0032] Optionally, the evaluation network inputs the calculated advantage function value into the strategy network, and the strategy network adjusts the control parameter according to the advantage function value, and the formula is as follows:

[0033]

[0034] Among them, θ t+1 represents the strategy parameter at time t+1, which is used to control the shape and characteristics of the control parameter a(t) at time t+1; θ t represents the strategy parameter at time t; α actor represents the learning rate of the policy network; represents the policy gradient, represents the probability of adjusting the policy parameters, π θ (a(t)|s(t)) means that under the environment state vector s(t), the policy function π θ The probability distribution of the chosen control parameter a(t) is where A(t) is the advantage function value.

[0035] As described above, an insecticide device and method provided by an embodiment of the present invention has the following beneficial effects: comprising a base, a support assembly, a shell, a collecting plate, an electronic control module and a sensor module: the base is movably connected to the shell through the support assembly; the shell is provided with an insect attracting assembly and an insecticide assembly, and the shell includes a concave cavity for forming a high-temperature cavity during operation; the collecting plate is arranged on the support assembly for collecting insects; the sensor module includes a temperature and humidity sensor, a light sensor, a wind speed sensor and a camera, and the temperature sensor, the light sensor, the wind speed sensor and the camera are communicatively connected with the electronic control module; the temperature and humidity sensor is used to collect the temperature and humidity of the environment, the light sensor is used to collect the light intensity of the environment, the wind speed sensor is used to collect the wind speed of the environment, and the camera is used to collect the image of the collecting plate; the electronic control module adjusts the control parameters of the insect attracting assembly and the insecticide assembly according to the collected temperature, humidity, light intensity, wind speed and image. The present invention combines environmental perception technology, multi-light source combination and adaptive adjustment algorithm, which can not only maintain excellent pest control effects in various complex environments, but also significantly reduce energy consumption, reduce adverse effects on the ecological environment, effectively improve insect control efficiency and have high environmental adaptability. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] Figure 1 It is a structural schematic diagram of an insect extermination device provided by an embodiment of the present invention;

[0037] Figure 2 It is a schematic diagram of a process of an insecticide elimination method provided by an embodiment of the present invention.

[0038] As shown in the figure: 1. base, 21. support rod, 22. bracket, 23. rotating shaft, 3. shell, 4. insect attracting component, 5. insect killing component, 6. collecting plate, 71. temperature and humidity sensor, 72. light sensor, 73. wind speed sensor. DETAILED DESCRIPTION

[0039] In order to enable those skilled in the art to better understand the technical solutions in the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.

[0040] See also Figure 1 to Figure 2It should be noted that the illustrations provided in this embodiment are only used to illustrate the basic concept of the present invention in a schematic manner, and the illustrations only show components related to the present invention rather than being drawn according to the number, shape and size of components in actual implementation. In actual implementation, the type, quantity and proportion of each component may be changed arbitrarily, and the component layout may also be more complicated.

[0041] See also Figure 1 , is a schematic diagram of the structure of an insecticide device provided by an embodiment of the present invention, such as Figure 1 As shown, the device includes a base 1, a support assembly 2 (not shown), a housing 3, a collection plate 6, an electronic control module 8 (not shown) and a sensor module 7 (not shown).

[0042] The base 1 is movably connected to the shell 3 via the support assembly 2 .

[0043] In an exemplary embodiment, the support assembly 2 may further include a support rod 21 and a bracket 22, wherein one end of the support rod 21 is arranged on the top surface of the base 1 and the other end is connected to the bracket 22; the bracket 22 is movably connected to the shell 3 through a rotating shaft 23, so that the shell 3 can rotate an angle during use, thereby being more conducive to capturing pests.

[0044] The shell 3 is provided with an insect attracting component 4 and an insect killing component 5. The shell 3 includes a concave cavity, which forms a high-temperature cavity during operation to kill the pests at high temperature.

[0045] In an exemplary embodiment, the insect attracting component 4 includes a high-pressure mercury lamp, which is arranged at the center of the inner wall of the cavity; the insect killing component 5 includes an iodine tungsten lamp, which is arranged at the edge of the inner wall of the cavity. Figure 1 Four insect-killing components 5 are exemplarily provided. In the specific implementation, the number of insect-killing components is not limited. Through the above configuration, the high-pressure mercury lamp is responsible for emitting a spectrum of ultraviolet band of 280-450nm, which can effectively attract target pests, while the iodine tungsten lamp emits infrared light of 500-1100nm and generates a high temperature of 500-800℃ on the lamp surface, forming a high-temperature killing area. Pests are lured by the high-pressure mercury lamp and fly into the high-temperature cavity formed by the iodine tungsten lamp and the concave lamp shell, and are quickly killed in the high-temperature environment.

[0046] Moreover, it should be noted that the insecticide device provided in the embodiment of the present invention adopts a modular design, that is, the base, support assembly, housing, insect attracting assembly and insecticide assembly can be flexibly disassembled and assembled according to actual application requirements. This modular structure facilitates the transportation, installation and maintenance of the equipment, and improves the overall convenience of use. At the same time, the annular arrangement of the high-pressure mercury lamp and the iodine tungsten lamp forms a three-dimensional insect attracting effect, which can cover a larger killing area. The design of the concave housing optimizes the focusing effect of light and heat, so that the light source (i.e., the high-pressure mercury lamp) and the heat source (i.e., the iodine tungsten lamp) work together to form an efficient killing area, thereby greatly improving the efficiency of insecticide removal.

[0047] The collecting plate 6 is arranged on the supporting assembly and is used to collect the killed insects. Specifically, the collecting plate 6 can be fixed on the supporting rod 21, and the insects killed by the high temperature chamber fall on the collecting plate 6 to realize the collection function.

[0048] The sensor module 7 includes a temperature and humidity sensor 71, a light sensor 72, a wind speed sensor 73 and a camera 74. The temperature and humidity sensor 71 is used to collect the temperature and humidity of the environment, the light sensor 72 is used to collect the light intensity of the environment, the wind speed sensor 73 is used to collect the wind speed of the environment, and the camera 74 is used to collect the image of the collection plate 6.

[0049] In an exemplary embodiment, the temperature and humidity sensor 71 and the light sensor 72 can be arranged on the top surface of the base 1, the wind speed sensor 73 can be arranged on the housing, and the camera 74 is arranged on the support assembly 2 and located above the collection plate 6 to collect the image on the collection plate 6. Optionally, the camera can be flexibly adjusted to different positions according to different device parameter models selected, but as long as the solution is arranged on the support assembly 2 and located above the collection plate 6, it should belong to the protection scope of this application.

[0050] The electronic control module 8 is connected in communication with the temperature and humidity sensor 71, the light sensor 72, the wind speed sensor 73 and the camera 74, so that the sensor module 7 collects data of the external environment in real time and transmits it to the electronic control module 8. The electronic control module 8 is further connected in communication with the insect attracting component 4 and the insect killing component 5, and is used to issue control instructions to the insect attracting component 4 and the insect killing component 5 according to the collected temperature, humidity, light intensity, wind speed and image, so as to adjust the control parameters of the insect attracting component 4 and the insect killing component 5, for example, when the insect attracting component 4 includes a high-pressure mercury lamp and the insect killing component includes an iodine tungsten lamp, control the light source intensity, control the temperature of the iodine tungsten lamp, control the high-pressure mercury lamp in a suitable light source band, etc. Through this refined control, the insect killing device can maintain the optimal operating state under various environmental conditions.

[0051] In specific implementation, the electronic control module 8 can be configured inside the base 1. Of course, in order to maintain the operation of the insecticide, other hardware components or peripheral circuits such as a power module, a communication module, and a storage module are also included, which will not be repeated in the embodiments of the present invention.

[0052] It can be seen from the description of the above embodiments that an insect extermination device provided by an embodiment of the present invention includes a base, a support assembly, a shell, a collecting plate, an electronic control module and a sensor module: the base is movably connected to the shell through the support assembly; the shell is provided with an insect attracting assembly and an insect extermination assembly, and the shell includes a concave cavity for forming a high-temperature cavity during operation; the collecting plate is arranged on the support assembly for collecting insects; the sensor module includes a temperature and humidity sensor, a light sensor, a wind speed sensor and a camera, and the temperature and humidity sensor, the light sensor, the wind speed sensor and the camera are communicatively connected with the electronic control module; the temperature and humidity sensor is used to collect the temperature and humidity of the environment, the light sensor is used to collect the light intensity of the environment, the wind speed sensor is used to collect the wind speed of the environment, and the camera is used to collect the image of the collecting plate; the electronic control module is communicatively connected with the insect attracting assembly and the insect extermination assembly, and adjusts the control parameters of the insect attracting assembly and the insect extermination assembly according to the collected temperature, humidity, light intensity, wind speed and image. The present invention adaptively adjusts the insecticidal lamp device for the light source of environmental perception, and realizes adaptive adjustment of the equipment environment in combination with the reinforcement learning algorithm, so as to effectively overcome the problems of poor environmental adaptability, high energy consumption and large impact on non-target organisms in the existing insecticidal technology. This device is based on the combination technology of multiple light sources such as high-pressure mercury lamps and iodine tungsten lamps, emitting light of different bands, greatly improving the insecticidal effect. At the same time, relying on intelligent algorithms, the device can dynamically perceive external environmental conditions (such as temperature, humidity, light intensity and wind speed, etc.), and adjust the light source intensity, heat source temperature and light source band selection in real time accordingly, so as to ensure that the device can maintain the optimal working state under various environmental conditions. In addition, the present invention also significantly improves ecological safety, and reduces interference with eco-friendly organisms such as beneficial insects by intelligently controlling the impact on non-target organisms, and further achieves energy saving and environmental protection goals. This technical solution combining intelligent algorithms with multiple light source combinations enables the device to have a high degree of adaptability and efficient insecticidal performance.

[0053] based on Figure 1The pest control device shown in the embodiment of the present invention also provides a pest control method, which is based on an intelligent control system of reinforcement learning to cope with complex environments. The main goal of this method is to achieve the best pest control effect under dynamic environmental conditions by automatically adjusting the control parameters of the device (such as light source intensity, heat source temperature, light source band selection, etc.), while maximizing energy conservation and reducing the impact on non-target organisms. In the specific implementation, it includes: environmental perception and state acquisition: obtaining the environmental state s(t)=[T(t), H(t), I(t), V(t)] in real time through temperature and humidity sensors, light sensors and wind speed sensors, and using it as the input of the reinforcement learning algorithm. Intelligent control system: Based on the Actor-Critic reinforcement learning algorithm, the intelligent control system dynamically selects the control parameters of the device (light source intensity, heat source temperature, light source band selection) according to the current environmental state s(t), and optimizes the performance of the device by calculating the reward function. Equipment operation and feedback: According to the control parameters output by the system, the device adjusts the working state of the light source and heat source, and obtains feedback information such as pest control effect, energy consumption and non-target biological impact, and calculates the reward value. Continuous optimization: The value of the current state is estimated through the Critic network, the advantage function is calculated and the Actor network is updated, so as to gradually optimize the control strategy in subsequent operations.

[0054] See also Figure 2 , is a flow chart of an insecticide method provided by an embodiment of the present invention. As shown in the figure, the execution subject of the method is the electronic control module 8 in the above embodiment. The steps of executing the insecticide method by the electronic control module 8 are as follows:

[0055] Step S101: obtain temperature, humidity, light and wind speed through temperature and humidity sensors, light sensors and wind speed sensors, and form an environmental state vector.

[0056] The pest control method provided by the present invention first relies on the environmental information collected by the sensor module, and uses these data as input states. Specifically, the electronic control module 8 obtains temperature and humidity through the temperature and humidity sensor 71, obtains light intensity through the light sensor 72, and obtains wind speed through the wind speed sensor, and further forms the environmental state vector with the obtained environmental information. The environmental state vector s(t) is a vector composed of multiple variables, expressed as: s(t) = [T(t), H(t), I(t), V(t)]. Wherein, T(t): the temperature sensor measures the real-time temperature of the environment where the pest control device is located (the temperature at time t); H(t): the humidity sensor measures the real-time humidity of the environment where the pest control device is located (the humidity at time t); I(t): the light sensor measures the real-time light intensity of the location of the pest control device (the light intensity or light intensity at time t); V(): the wind speed sensor measures the real-time wind speed of the environment where the pest control device is located (the wind speed at time t). This environmental information may affect the pest control effect, and these environmental data are transmitted to the electronic control module 8 in real time for calculating the subsequent decision-making process.

[0057] Step S102: Input the environmental state vector into the strategy network to calculate control parameters, wherein the control parameters include the strength of the insect attracting component and the insect killing component, the temperature of the insect killing component, and the waveband of the insect attracting component.

[0058] Based on the current environment state vector, the Actor network (i.e., the policy network) decides the next action, which is to adjust the control parameters of the insect-killing device (i.e., adjust the intensity of the insect-attracting component and the insect-killing component, the temperature of the insect-killing component, and the wavelength of the insect-attracting component). The control parameters a(t) to be adjusted include the intensity P light (t), temperature T heat (t) and band B(t), which represents the control action of the insecticide at time t. Specifically, the output of the Actor network is:

[0059] a(t)=σ(MLP(s(t)))=[P light (t),T heat (t),B(t)]

[0060] Where MLP represents full connection; σ represents the Sigmoid activation function, which limits the output to a specific range (such as light intensity percentage and temperature range) to obtain the parameters to be controlled; P light (t) Control intensity (when the insect attracting component is a high-pressure mercury lamp and the insect killing component is an iodine tungsten lamp, the intensity is the light source intensity): control the light intensity output of the high-pressure mercury lamp and the iodine tungsten lamp to adapt to different environmental requirements; T heat(t) Control temperature: When the insect killing component is an iodine tungsten lamp, control the temperature of the iodine tungsten lamp heat source to ensure the killing effect when the humidity is high; B (t) Select the appropriate band: When the insect attractant is a high-pressure mercury lamp, adjust the band combination of the high-pressure mercury lamp light source to maximize the pest capture rate and reduce the impact on non-target organisms. The electronic control module 8 uses a neural network to comprehensively evaluate the collected environmental data and generate a control parameter combination that best suits the current environmental conditions. Moreover, the electronic control module 8 not only adjusts each control parameter individually, but also considers the interdependence between them. For example, in a high-light and high-wind speed environment, the electronic control module 8 will simultaneously reduce the intensity of the light source (i.e., the high-pressure mercury lamp and the iodine tungsten lamp), increase the temperature of the heat source (i.e., the iodine tungsten lamp), and select a light source with a wider band (i.e., the high-pressure mercury lamp) to ensure that the insect killing effect is not affected by changes in light and wind speed.

[0061] Step S103: Calculate the reward value by weighted calculation based on the capture rate, energy consumption and ecological impact index, wherein the capture rate is the ratio of the number of target organisms captured per unit time to the total capture amount obtained by analyzing the images collected by the camera, the energy consumption is the energy consumption of the insecticide, and the ecological impact index is the ratio of the number of non-target organisms captured to the total capture amount.

[0062] The electronic control module adjusts the operating state according to the control parameters calculated by the strategy network and performs the corresponding actions. After execution, the system obtains the following feedback. Taking into account the capture rate, energy consumption, ecological impact index and other indicators, the reward value is obtained through weighted calculation. The reward function for calculating the reward value is as follows:

[0063] R(t) = α·capture rate(t)-β·energy consumption(t)-γ·ecological impact index(t)

[0064] Among them: the capture rate (t) represents the capture efficiency of the insecticide device on the target organism at time t in the current environment. The acquisition method is to use the camera installed on the support component to monitor and record the number of captured pests (i.e., target organisms) and beneficial insects (i.e., non-target organisms) in real time, and then calculate the number of pests captured per unit time, and then obtain the capture rate. Energy consumption (t) represents the energy consumption of the insecticide device at the current moment t. It monitors the current consumption of each power-consuming component of the device, such as the light source and the heat source, in real time through current detection, and calculates the actual power consumption of each power-consuming component, such as the light source and the heat source, in combination with the voltage data, and summarizes the overall energy consumption. The ecological impact index (t) represents the impact of the insecticide device at the current moment t on non-target organisms (such as beneficial insects). The acquisition method is to use a camera to monitor non-target organisms (such as beneficial insects), evaluate the number of non-target organisms, and calculate the ecological impact index. In an exemplary embodiment, the unit time can be set to 10 minutes, so that the number of captured pests and 1 beneficial insect is obtained every 10 minutes, and the capture rate (t) is calculated to be 15 / 16×100%=93.75%, and its ecological impact index (t)=1 / 16×100%=6.25%. In addition, it should be noted that for the identification of pests and beneficial insects, in the specific implementation, the open source model yolov8 can be used to train the data set of pests and beneficial insects, so as to obtain the types of captured insects identified by the pest and beneficial insect discriminator, which will not be repeated in the embodiments of the present invention.

[0065] The weight parameters α, β, and γ are used to balance capture efficiency, energy consumption, and ecological safety, where the above weight parameters satisfy the following relationship:

[0066] α+β+γ=1, 0<α,β,γ<1, α∈[0.4,0.6], β∈[0.3,0.5], γ∈[0.1,0.3]

[0067] α is the capture efficiency weight, which reflects the ability of the pest control device to capture target pests. A higher α value means that more attention is paid to improving the capture rate. β is the energy consumption weight, which reflects the energy consumption of pest control. A higher β value means that more attention is paid to saving energy and reducing energy consumption. γ is the ecological impact weight, which reflects the impact of the pest control device on non-target organisms. A higher γ value means that more attention is paid to reducing the impact on eco-friendly organisms such as beneficial insects. The goal of the entire pest control device is to maximize the reward function R(t), that is, to reduce energy consumption and reduce the impact on non-target organisms while ensuring efficient capture of pests.

[0068] Step S104: Calculate the advantage function value according to the calculated reward value and the value estimate corresponding to the environment state vector at different times, wherein the value estimate is a scalar value calculated by the evaluation network through a multi-layer neural network according to the input environment state vector at different times.

[0069] The evaluation network of the insecticide device uses a Critic network in the embodiment of the present invention to evaluate the current environmental state. The Critic network estimates the value V of the current environmental state vector s(t) φ (s(t)) = f Critic (s(t)), where f Critic The network consists of a multi-layer feedforward neural network, including an input layer, multiple hidden layers, and an output layer. The input layer receives the environment state vector s(t), performs nonlinear transformation through the hidden layer, and finally outputs a scalar value representing the value estimate V of the environment state vector φ (s(t)). Similarly, after obtaining the new environment state vector s(t+1), that is, the environment state vector at time t+1, the Critic network estimates the value estimate V of the new environment state vector s(t+1) in the manner described in the above embodiment. φ (s(t+1))=f Critic (s(t+1)). The operating performance of the pest control device under these environmental conditions directly affects the final pest control efficiency, energy consumption, and impact on non-target organisms.

[0070] In order to further improve the performance of the insect-killing device, the Critic network calculates the advantage function by comparing the actual reward with the expected value, and obtains the advantage function value A(t) to guide the Actor network (i.e., the policy network) to optimize the control strategy. The calculation formula of the advantage function value A(t) is as follows:

[0071] A(t)=R(t)+δ·V φ (s(t+1))-V φ (s(t))

[0072] Where R(t) is the reward value calculated at time t, including capture rate, energy consumption, and impact on non-target organisms, δ is the discount factor, 0≤δ<1, which is used to balance short-term and long-term rewards, V φ (s(t+1)) and V φ (s(t)) are the value estimates of the current and new environmental state vectors, respectively. In this way, the Critic network can estimate the "goodness" of the current state of the insecticide. Specifically, the Critic network evaluates the "goodness" of the current state and action through the positive and negative values ​​and size of the advantage function value A(t). The larger the advantage function value, the better the performance of the current action in this state. The Critic network considers this state to be "good". The smaller or negative the advantage function value, the poorer the performance of the current action in this state. The Critic network considers this state to be "bad" and provides the advantage function value feedback to the Actor network to improve the control strategy.

[0073] Step S105: the evaluation network inputs the calculated advantage function value into the strategy network, and the strategy network adjusts the control parameter according to the advantage function value.

[0074] In the embodiment of the present invention, the Actor network is responsible for selecting appropriate control parameters (such as the light source intensity, heat source temperature, light source band, etc. described in the above embodiment) according to the current environmental state vector, and continuously optimizing its control strategy using the policy gradient method. Specifically, the Actor network estimates the current environmental state and decides how to adjust the various control parameters of the insecticide device to maximize the reward function value R(t) and keep the device performing optimally under environmental changes. The control parameter update formula of the Actor network is as follows:

[0075]

[0076] Among them, α actor is the learning rate of the Actor network, which is configured to be 0.0001 in this embodiment of the present invention. is the policy gradient, where the policy function π θ (a(t)|s(t)) means that under the environment state vector s(t), the policy function π θ The policy parameter θ controls the shape and properties of this probability distribution. It is used to indicate how to adjust the parameter θ to increase or decrease the probability of a specific control parameter, thereby optimizing the strategy. The strategy function indicates how to update the control strategy to adapt to the current environment. A(t) is the advantage function value (i.e., the calculation result of the advantage function), which is used to measure the current control light intensity P light (t), heat source temperature T heat (t) and the light source band selection B(t) to achieve the insect killing effect. With the update process of this learned control strategy, the Actor network can gradually achieve dynamic optimization to ensure that the device can achieve the best effect under various environmental conditions.

[0077] Of course, it should be noted that the steps described in the above embodiments are continuously repeated and iterated, so that the control strategy can be continuously optimized to ensure that the insecticide device maintains the best insecticide effect and energy efficiency in a dynamic environment.

[0078] In the pest control method provided by the embodiment of the present invention, the Critic network is responsible for evaluating the performance of the device under the current environmental state. Specifically, the Critic network evaluates the pest control effect, energy efficiency and impact on non-target organisms of the pest control device under specific working conditions by calculating the "value" (i.e., value estimation) of the current environmental state. Based on this, the Critic network can generate a feedback signal (i.e., advantage function value) to reflect the performance of the current control strategy, and provide this feedback signal to the Actor network. The Actor network uses this feedback information to adjust the control strategy (control parameters such as light source intensity, heat source temperature, light source band selection, etc.), thereby optimizing the control parameters of the pest control device, improving pest control efficiency, reducing energy consumption, and minimizing the impact on non-target organisms. In this way, the Critic network and the Actor network cooperate with each other to form a closed-loop control system to ensure that the pest control device can adjust the working parameters in real time according to the dynamic changes of the environment, so as to always maintain the optimal working state.

[0079] The coordinated adjustment of light source intensity, heat source temperature and light source band selection is optimized through the collaboration of the Critic network and the Actor network to ensure the coordinated work of various parameters. For example, in an environment of high temperature, high humidity and low wind speed, the insect killer will simultaneously increase the heat source temperature, maintain a high light source intensity, and select a concentrated band of light sources to ensure efficient killing of pests. In an environment of low temperature, low humidity and high wind speed, the insect killer will lower the heat source temperature, lower the light source intensity, and select a wide band of light sources to save energy while expanding the trapping range. Through this control strategy of multi-light source combination and parameter coordinated optimization, the insect killer can achieve the best insect killing effect and energy efficiency balance under different environmental conditions.

[0080] From the description courseware of the above embodiments, the pest control method provided by the present invention based on the pest control device in the above embodiments achieves efficient pest control effects under different environmental conditions through multi-light source adjustment and intelligent control technology, and provides an environmentally friendly alternative for agricultural pest control. The device uses a combination of high-pressure mercury lamps and iodine tungsten lamps, combined with an environmental perception system and an adaptive adjustment algorithm, to dynamically adjust the light source intensity, band selection and heat source temperature to adapt to environmental changes such as different temperatures, humidity, light intensity and wind speed, ensuring that the pest control efficiency is always in the best state.

[0081] The intelligent control system of the insecticide device integrates an environmental perception sensor module, including temperature and humidity, light intensity and wind speed sensors, which can collect external environmental conditions in real time and transmit them to the adaptive adjustment algorithm. The insecticide method dynamically adjusts the light source intensity, band combination and heat source temperature of the high-pressure mercury lamp and the iodine tungsten lamp based on the collected data. In actual use, for low-light environments, increasing the light source intensity of the high-pressure mercury lamp (such as from 50% to 100%) and selecting a wider band (300-600 nanometers) can attract pests in a larger range and increase the capture rate. In high-light environments, reducing the light source intensity (such as from 100% to 50%) and selecting a more concentrated band (350-450 nanometers) can avoid energy waste while maintaining the trapping effect. In high-temperature and high-humidity environments, increasing the heat source temperature of the iodine tungsten lamp (such as from 500°C to 750°C) needs to be coordinated with appropriate light source intensity (such as maintaining high light source intensity) to ensure that high temperature can quickly kill pests without causing excessive energy consumption. In low temperature and low humidity environments, reducing the temperature of the heat source (e.g., from 750°C to 500°C) combined with reducing the intensity of the light source (e.g., from 100% to 50%) can save energy while still maintaining a basic killing effect. Experiments have shown that the adaptive light source adjustment device based on environmental perception has a significant attraction to common agricultural pests (such as aphids, cotton bollworms, fall armyworms, etc.), with a capture efficiency of more than 95%, and significantly improved the pest control effect.

[0082] To sum up, the embodiments of the present invention realize an efficient, intelligent and environmentally friendly insect extermination device and an insect extermination method based on the insect extermination device by combining environmental perception technology, multi-light source combination and adaptive adjustment algorithm. It can not only maintain excellent pest control effects in various complex environments, but also significantly reduce energy consumption and reduce adverse effects on the ecological environment, providing an innovative solution for pest control fields such as agriculture, forestry, and aquaculture.

[0083] Through the description of the above method embodiments, the technicians in the relevant field can clearly understand that the present invention can be implemented by means of software plus the necessary general hardware platform, and of course it can also be implemented by hardware, but in many cases the former is a better implementation method. Based on such an understanding, the technical solution of the present invention can be essentially or partly embodied in the form of a software product that contributes to the prior art. The computer software product is stored in a storage medium and includes several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, and other media that can store program codes.

[0084] An embodiment of the present invention provides a non-volatile computer storage medium, wherein the computer storage medium stores computer executable instructions, and the computer executable instructions can execute the pest control method in any of the above method embodiments.

[0085] The above-mentioned pest control method is based on the pest control device provided by the embodiment of the present invention, and has the corresponding functional modules and beneficial effects of the pest control device. For technical details not described in detail in this embodiment, please refer to the pest control device provided by the embodiment of the present invention.

[0086] The device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, i.e., they may be located in one place or distributed over multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the present embodiment.

[0087] Each embodiment in this specification is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the device or system embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment. The device and system embodiments described above are merely schematic, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Ordinary technicians in this field can understand and implement it without paying creative labor.

[0088] It should be noted that, in this article, relational terms such as "first" and "second" are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the sentence "comprise a ..." do not exclude the existence of other identical elements in the process, method, article or device including the elements.

[0089] The above description is only a specific embodiment of the present invention, so that those skilled in the art can understand or implement the present invention. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but should conform to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. An insect extermination device, characterized in that: Includes base, support assembly, housing, collection plate, electronic control module and sensor module: The base is movably connected to the shell through a support assembly; The housing is provided with an insect attracting component and an insect killing component, and the housing includes a concave cavity for forming a high-temperature cavity during operation; The collecting plate is arranged on the supporting assembly and is used to collect insects; The sensor module includes a temperature and humidity sensor, a light sensor, a wind speed sensor and a camera; the temperature and humidity sensor, the light sensor, the wind speed sensor and the camera are communicatively connected with the electronic control module; the temperature and humidity sensor is used to collect the temperature and humidity of the environment, the light sensor is used to collect the light intensity of the environment, the wind speed sensor is used to collect the wind speed of the environment, and the camera is used to collect the image of the collection plate; the electronic control module is communicatively connected with the insect attracting component and the insect killing component, and the control parameters of the insect attracting component and the insect killing component are calculated and adjusted through a reinforcement learning algorithm according to the collected temperature, humidity, light intensity, wind speed and image.

2. The insect extermination device according to claim 1, characterized in that: The temperature and humidity sensor and the light sensor are arranged on the top surface of the base, the wind speed sensor is arranged on the shell, and the camera is arranged above the support component and the collecting plate to collect the image on the collecting plate.

3. The insect extermination device according to claim 1, characterized in that: The insect attracting component comprises a high-pressure mercury lamp, which is arranged at the center of the inner wall of the concave cavity; the insect killing component comprises an iodine tungsten lamp, which is arranged at the edge of the inner wall of the concave cavity.

4. The insecticide device according to claim 1, characterized in that: The support assembly includes a support rod and a bracket, one end of the support rod is arranged on the top surface of the base, and the other end is connected to the bracket; the bracket is movably connected to the shell through a rotating shaft so that the shell can rotate at an angle.

5. An insect extermination method based on the insect extermination device according to claims 1-4, characterized in that: include: The temperature, humidity, light intensity and wind speed are obtained through the temperature and humidity sensor, light sensor and wind speed sensor, and form the environmental state vector; Input the environmental state vector into the strategy network to calculate the control parameters, wherein the control parameters include the strength of the insect attracting component and the insect killing component, the temperature of the insect killing component, and the band of the insect attracting component; The reward value is calculated by weighted calculation based on the capture rate, energy consumption and ecological impact index, wherein the capture rate is the ratio of the number of target organisms captured per unit time to the total capture obtained by analyzing the images collected by the camera, the energy consumption is the energy consumption of the insecticide, and the ecological impact index is the ratio of the number of non-target organisms captured to the total capture; Calculate the advantage function value according to the calculated reward value and the value estimate corresponding to the environment state vector at different times, wherein the value estimate is a scalar value calculated by the evaluation network through a multi-layer neural network according to the input environment state vector at different times; The evaluation network inputs the calculated advantage function value into the strategy network, and the strategy network adjusts the control parameter according to the advantage function value.

6. The method for disinsection according to claim 5, characterized in that: The environment state vector is input into the strategy network to calculate the control parameters, and the formula is as follows: a(t)=σ(MLP(s(t)))=[P light (t),T heat (t),B(t)] Among them, a(t) represents the control parameter, σ represents the Sigmoid activation function, MLP represents full connection, s(t) represents the environment state vector at time t, P light (t) represents the calculated intensity at time t, T heat (t) represents the temperature at time t, B(t) represents the band at time t, and the strategy network includes an Actor network.

7. The insecticide method according to claim 5, characterized in that: The reward value is calculated by weighting according to the capture rate, energy consumption and ecological impact index, and the formula is as follows: R(t) = α·capture rate(t)-β·energy consumption(t)-γ·ecological impact index(t) Among them, R(t) represents the reward value at time t, capture rate(t) represents the capture rate at time t, energy consumption(t) represents the energy consumption at time t, ecological impact index(t) represents the ecological impact index at time t, α represents the capture rate weight, β represents the energy consumption weight, and γ represents the ecological impact weight.

8. The insecticide method according to claim 7, characterized in that: The capture rate weight, the energy consumption weight and the ecological impact weight are as follows: α+β+γ=1 0<α, β, γ<1 α∈[0.4,0.6], β∈[0.3,0.5], γ∈[0.1,0.3].

9. The insecticide method according to claim 5, characterized in that: The advantage function value is calculated based on the calculated reward value and the value estimate corresponding to the environment state vector at different times. The formula is as follows: A(t)=R(t)+δ·V φ (s(t+1))-V φ (s(t)) Among them, A(t) represents the advantage function value at time t, R(t) represents the reward value at time t, and V φ (s(t+1)) represents the value estimate at time t+1, s(t+1) represents the environment state vector at time t+1, V φ (s(t)) represents the value estimate at time t, s(t) represents the environment state vector at time t, δ is the discount factor, and 0≤δ<1, and the evaluation network includes a Critic network.

10. The insecticide method according to claim 5, characterized in that: The evaluation network inputs the calculated advantage function value into the strategy network, and the strategy network adjusts the control parameter according to the advantage function value. The formula is as follows: Among them, θ t+1 represents the strategy parameter at time t+1, which is used to control the shape and characteristics of the control parameter a(t) at time t+1; θ t represents the strategy parameter at time t; α actor represents the learning rate of the policy network; represents the policy gradient, represents the probability of adjusting the policy parameters, π θ (a(t)|s(t)) means that under the environment state vector s(t), the policy function π θ The probability distribution of the chosen control parameter a(t) is where A(t) is the advantage function value.

Images