Insect killing device and method

Abstract

An insect killing device and method. The device comprises a base (1), a support assembly (2), a housing (3), a collection plate (6), an electronic control module (8) and a sensor module (7). The base (1) is movably connected to the housing (3) by means of the support assembly (21); the housing (3) is provided with an insect-attracting assembly (4) and insect-killing assemblies (5), and the housing (3) comprises an recessed cavity which forms a high-temperature cavity during working; the collection plate (6) is provided on the support assembly (2); the sensor module (7) comprises a temperature and humidity sensor (71), a light intensity sensor (72), a wind speed sensor (73) and a camera (74), all of which are in communication connection with the electronic control module (8); and the electronic control module (8) adjusts control parameters of the insect-attracting assembly (4) and the insect-killing assemblies (5) on the basis of the collected temperature, humidity, light intensity, wind speed and images. In combination with the environment sensing technology, multi-light source combination and an adaptive adjustment algorithm, the device not only can maintain excellent pest control effects in various complex environments, but also significantly reduce energy consumption and reduce adverse effects on the ecological environment, thus effectively improving insect killing efficiency and having high environmental adaptability.

Description

An insect control device and method Technical Field

[0001] This invention relates to the field of agricultural pest control, specifically to an insecticidal device and method. Background Technology

[0002] Agricultural pest control is a crucial task in agricultural production, affecting crop yield and quality, and consequently directly impacting national food security and farmers' income. Traditional pest control methods primarily rely on chemical pesticides. While effective in the short term, long-term use not only increases pest resistance but also severely impacts the environment and human health. In recent years, with increased environmental awareness and technological advancements, finding efficient and environmentally friendly pest control methods has become a research hotspot. Among these, high-temperature sterilization technology based on light sources has gained increasing attention due to its advantages of being pollution-free and energy-efficient. Existing technologies are mostly based on fixed combinations of light and heat sources, but in practical applications, the effectiveness of these devices often falls short of optimization due to varying environmental conditions. Different environmental factors, such as temperature, humidity, light intensity, and wind speed, all affect the efficiency and energy consumption of pest control devices.

[0003] As can be seen from the above description, current agricultural pest control is difficult to adapt flexibly to the environment. How to improve pest control efficiency and environmental adaptability is a technical problem that urgently needs to be solved by those skilled in the art. Summary of the Invention

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

[0005] To achieve the above objectives, 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 housing via a support assembly;

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

[0008] The collection plate is mounted on the support 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. These sensors are communicatively connected to the electronic control module. The temperature and humidity sensor collects ambient temperature and humidity data, the light sensor collects ambient light intensity data, the wind speed sensor collects ambient wind speed data, and the camera collects images of the collection plate. The electronic control module is communicatively connected to the insect-attracting and insect-killing components. Based on the collected temperature, humidity, light intensity, wind speed, and images, it uses a reinforcement learning algorithm to calculate and adjust the control parameters of the insect-attracting and insect-killing components.

[0010] Optionally, the temperature and humidity sensor and the light sensor are disposed on the top surface of the base, the wind speed sensor is disposed on the housing, and the camera is disposed above the support assembly and the collection plate for collecting images on the collection plate.

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

[0012] Optionally, the support assembly includes a support rod and a bracket, with one end of the support rod disposed on the top surface of the base and the other end connected to the bracket; the bracket is movably connected to the housing via a rotating shaft, allowing the housing to rotate by an angle.

[0013] According to a second aspect of the present invention, embodiments of the present invention also provide an insecticidal method based on the insecticidal device described above, the method comprising:

[0014] Temperature, humidity, light intensity, and wind speed are acquired through temperature and humidity sensors, light sensors, and wind speed sensors, and then combined to form an environmental state vector.

[0015] The environmental state vector is input into the policy network to calculate the control parameters, which include 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.

[0016] The reward value is calculated by weighting the capture rate, energy consumption, and ecological impact index. The capture rate is the ratio of the number of target organisms captured per unit time to the total number of organisms captured, obtained by analyzing images collected by the camera. The energy consumption is the energy consumption of the insecticidal device. The ecological impact index is the ratio of the number of non-target organisms captured to the total number of organisms captured.

[0017] Based on the calculated reward value and the value estimate corresponding to the environmental state vector at different times, the advantage function value is calculated, wherein the value estimate is a scalar value calculated by the evaluation network through a multi-layer neural network based on the input environmental state vector at different times.

[0018] The evaluation network inputs the calculated advantage function value into the policy network, and the policy network adjusts the control parameters based on the advantage function value.

[0019] Optionally, the environmental state vector is input into the policy network to calculate the control parameters, as follows: a(t)=σ(MLP(s(t)))=[P light (t),T heat (t),B(t)]

[0020] Where a(t) represents the control parameter, σ represents the Sigmoid activation function, MLP represents a fully connected layer, s(t) represents the environment state vector at time t, and P light (t) represents the calculated intensity at time t, T heat B(t) represents the temperature at time t, B(t) represents the band at time t, and the policy network includes an Actor network.

[0021] Optionally, the reward value is calculated by weighting the capture rate, energy consumption, and ecological impact index, as follows: R(t) = α·capture rate(t) - β·energy consumption(t) - γ·ecological impact index(t)

[0022] Where 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.

[0023] Optionally, the capture rate weight, the energy consumption weight, and the ecological impact weight are taken 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].

[0024] Optionally, the advantage function value is calculated based on the calculated reward value and the value estimate corresponding to the environmental state vector at different times, as shown in the following formula: A(t)=R(t)+δ·V φ (s(t+1))-V φ (s(t))

[0025] Where 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, and V φ (s(t)) represents the value estimate at time t, s(t) represents the environmental state vector at time t, δ is the discount factor, and 0≤δ<1. The evaluation network includes the Critic network.

[0026] Optionally, the evaluation network inputs the calculated advantage function value into the policy network, and the policy network adjusts the control parameters according to the advantage function value, as shown in the following formula:

[0027] Where, θ t+1 θ represents the policy parameters at time t+1, used to control the shape and characteristics of the selection control parameter a(t) at time t+1; t Represents the policy parameters at time t; α actor This represents the learning rate of the policy network; Represents the policy gradient. π represents the probability of adjusting the policy parameters. θ (a(t)|s(t)) represents the policy function π given the environment state vector s(t). θ Choose the probability distribution of the control parameter a(t), where A(t) is the dominance function value.

[0028] As described above, the insect-killing device and method provided by the embodiments of the present invention have the following beneficial effects: The device includes a base, a support assembly, a housing, a collection plate, an electronic control module, and a sensor module. The base is movably connected to the housing via the support assembly. The housing is provided with an insect-attracting component and an insect-killing component, and includes a cavity for forming a high-temperature chamber during operation. The collection plate is disposed 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. The temperature sensor, light sensor, wind speed sensor, and camera are communicatively connected to the electronic control module. The temperature and humidity sensor is used to collect the ambient temperature and humidity, the light sensor is used to collect the ambient light intensity, the wind speed sensor is used to collect the ambient wind speed, and the camera is used to collect images of the collection plate. The electronic control module adjusts the control parameters of the insect-attracting component and the insect-killing component based on the collected temperature, humidity, light intensity, wind speed, and images. This invention combines environmental sensing technology, multi-source light source combination, and adaptive adjustment algorithm, which can not only maintain excellent pest control effect in various complex environments, but also significantly reduce energy consumption, reduce adverse impact on the ecological environment, effectively improve pest control efficiency, and have high environmental adaptability. Attached Figure Description

[0029] Figure 1 is a schematic diagram of the structure of an insecticidal device provided in an embodiment of the present invention;

[0030] Figure 2 is a schematic flowchart of an insect control method provided in an embodiment of the present invention.

[0031] The following components are shown in the figure: 1. Base, 21. Support rod, 22. Bracket, 23. Rotating shaft, 3. Housing, 4. Insect attracting component, 5. Insect killing component, 6. Collection plate, 71. Temperature and humidity sensor, 72. Light sensor, 73. Wind speed sensor. Detailed Implementation

[0032] To enable those skilled in the art to better understand the technical solutions of this invention, the technical solutions of the embodiments of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this invention, and not all embodiments. Based on the embodiments of this invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of this invention.

[0033] Please refer to Figures 1 and 2. It should be noted that the illustrations provided in this embodiment are only schematic representations of the basic concept of the present invention. Therefore, the illustrations only show the components related to the present invention and are not drawn according to the actual number, shape and size of the components in the actual implementation. In the actual implementation, the form, quantity and proportion of each component can be arbitrarily changed, and the layout of the components may also be more complex.

[0034] Referring to Figure 1, it is a structural schematic diagram of an insect-killing device provided in an embodiment of the present invention. As shown in Figure 1, the device includes a base 1, a support component 2 (not shown), a housing 3, a collection plate 6, an electronic control module 8 (not shown), and a sensor module 7 (not shown).

[0035] The base 1 is movably connected to the shell 3 via the support component 2.

[0036] In an exemplary embodiment, the support assembly 2 may further include a support rod 21 and a bracket 22. One end of the support rod 21 is disposed 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 housing 3 via a rotating shaft 23, so that the housing 3 can rotate at an angle during use, thereby making it more convenient to capture pests.

[0037] The housing 3 is equipped with an insect-attracting component 4 and an insect-killing component 5. The housing 3 includes a concave cavity, which forms a high-temperature cavity during operation to kill pests with high temperature.

[0038] In an exemplary embodiment, the insect-attracting component 4 includes a high-pressure mercury lamp disposed at the center of the inner wall of the concave cavity; the insect-killing component 5 includes an iodine-tungsten lamp disposed at the edge of the inner wall of the concave cavity. Figure 1 exemplarily shows four insect-killing components 5, but the number of components is not limited in specific implementations. With this configuration, the high-pressure mercury lamp emits ultraviolet light in the 280-450nm band, effectively attracting target pests, while the iodine-tungsten lamp emits infrared light in the 500-1100nm band and generates a high temperature of 500-800°C on its surface, forming a high-temperature killing zone. Attracted by the high-pressure mercury lamp, pests fly into the high-temperature cavity formed by the iodine-tungsten lamp and the concave lamp housing, where they are rapidly killed in the high-temperature environment.

[0039] Furthermore, it should be noted that the insect-killing device provided in this embodiment of the invention adopts a modular design, meaning that the base, support components, shell, insect-attracting components, and insect-killing components can be flexibly disassembled and combined according to actual application needs. This modular structure facilitates the transportation, installation, and maintenance of the equipment, improving overall ease of use. Simultaneously, the circular arrangement of the high-pressure mercury lamp and the tungsten iodine lamp creates a three-dimensional insect-attracting effect, covering a larger killing area. The concave shell design optimizes the focusing effect of light and heat, allowing the light source (i.e., the high-pressure mercury lamp) and the heat source (i.e., the tungsten iodine lamp) to work synergistically, forming a highly efficient killing area, thereby significantly improving insect-killing efficiency.

[0040] The collecting plate 6 is mounted on the support assembly and is used to collect the killed insects. Specifically, the collecting plate 6 can be fixed to the support rod 21, and the insects killed by the high-temperature chamber fall onto the collecting plate 6, thus achieving the collection function.

[0041] 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 images of the collection plate 6.

[0042] In one exemplary embodiment, the temperature and humidity sensor 71 and the light sensor 72 can be disposed on the top surface of the base 1, the wind speed sensor 73 can be disposed on the housing, and the camera 74 is disposed on the support assembly 2 and located above the collection plate 6 to acquire images on the collection plate 6. Optionally, the camera can be flexibly adjusted to different positions depending on the selected device parameters and models, but any arrangement where the camera is disposed on the support assembly 2 and located above the collection plate 6 should fall within the protection scope of this application.

[0043] The electronic control module 8 is communicatively connected to the temperature and humidity sensor 71, the light sensor 72, the wind speed sensor 73, and the camera 74. This allows the sensor module 7 to collect real-time data about the external environment and transmit it to the electronic control module 8. The electronic control module 8 is also further communicatively connected to the insect-attracting component 4 and the insect-killing component 5. Based on the collected temperature, humidity, light intensity, wind speed, and images, the module sends control commands to the insect-attracting component 4 and the insect-killing component 5 to adjust their control parameters. For example, when the insect-attracting component 4 includes a high-pressure mercury lamp and the insect-killing component includes an iodine-tungsten lamp, the module controls the light source intensity, the temperature of the iodine-tungsten lamp, and the high-pressure mercury lamp to operate in a suitable wavelength range. Through this precise control, the insect-killing device can maintain optimal operation under various environmental conditions.

[0044] In a specific implementation, the electronic control module 8 can be configured inside the base 1. Of course, in order to maintain the operation of the insect-killing device, other hardware components or peripheral circuits such as power supply module, communication module, and storage module are also included, which will not be described in detail in the embodiments of the present invention.

[0045] As described in the above embodiments, the insect-killing device provided by the present invention includes a base, a support assembly, a housing, a collection plate, an electronic control module, and a sensor module. The base is movably connected to the housing via the support assembly. The housing is provided with an insect-attracting assembly and an insect-killing assembly, and the housing includes a cavity for forming a high-temperature chamber during operation. The collection plate is disposed 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. The temperature and humidity sensor, the light sensor, the wind speed sensor, and the camera are communicatively connected to the electronic control module. The temperature and humidity sensor is used to collect the ambient temperature and humidity, the light sensor is used to collect the ambient light intensity, the wind speed sensor is used to collect the ambient wind speed, and the camera is used to collect images of the collection plate. The electronic control module is communicatively connected to the insect-attracting assembly and the insect-killing assembly, and adjusts the control parameters of the insect-attracting assembly and the insect-killing assembly based on the collected temperature, humidity, light intensity, wind speed, and images. This invention relates to an environmentally sensitive, adaptively adjustable insecticidal lamp device. By incorporating reinforcement learning algorithms, it achieves adaptive adjustment of the device to its environment, effectively overcoming the problems of poor environmental adaptability, high energy consumption, and significant impact on non-target organisms in existing insecticidal technologies. The device utilizes a combination of multiple light sources, including high-pressure mercury lamps and tungsten iodine lamps, to emit light at different wavelengths, significantly improving insecticidal efficacy. Simultaneously, relying on intelligent algorithms, the device can dynamically sense external environmental conditions (such as temperature, humidity, light intensity, and wind speed) and adjust the light source intensity, heat source temperature, and light source wavelength selection in real time, ensuring optimal operation under various environmental conditions. Furthermore, this invention significantly improves ecological safety by intelligently controlling the impact on non-target organisms, reducing interference with beneficial insects and other eco-friendly organisms, further achieving energy conservation and environmental protection goals. This technical solution combining intelligent algorithms and multiple light sources gives the device high adaptability and efficient insecticidal performance.

[0046] Based on the insecticidal device shown in Figure 1, this embodiment of the invention also provides an insecticidal method, using a reinforcement learning-based intelligent control system to cope with complex environments. The main objective of this method is to achieve optimal insecticidal effect under dynamic environmental conditions by automatically adjusting the control parameters of the equipment (such as light source intensity, heat source temperature, and light source wavelength selection), while maximizing energy conservation and minimizing the impact on non-target organisms. In specific implementation, this includes: Environmental perception and state acquisition: Real-time acquisition of the environmental state s(t) = [T(t), H(t), I(t), V(t)] using temperature and humidity sensors, light sensors, and wind speed sensors, and using this as input to the reinforcement learning algorithm. Intelligent control system: Based on the Actor-Critic reinforcement learning algorithm, the intelligent control system dynamically selects the equipment's control parameters (light source intensity, heat source temperature, and light source wavelength selection) according to the current environmental state s(t), and optimizes the equipment's performance by calculating a reward function. Equipment operation and feedback: Based on the control parameters output by the system, the equipment adjusts the working state of the light source and heat source, while simultaneously acquiring feedback information such as insecticidal effect, energy consumption, and the impact on non-target organisms, and calculating 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, thereby gradually optimizing the control strategy in subsequent operations.

[0047] Referring to Figure 2, which is a flowchart illustrating a pest control method according to an embodiment of the present invention, as shown in the figure, the main body executing this method is the electronic control module 8 in the above embodiment. The steps of the pest control method executed by the electronic control module 8 are as follows:

[0048] Step S101: Obtain temperature, humidity, light intensity, and wind speed through temperature and humidity sensors, light sensors, and wind speed sensors, and form an environmental state vector.

[0049] The insect control method provided by this invention first relies on environmental information collected by a sensor module, using this data as input. Specifically, the electronic control module 8 acquires temperature and humidity through a temperature and humidity sensor 71, light intensity through a light sensor 72, and wind speed through a wind speed sensor. The acquired environmental information is further organized into an environmental state vector, s(t), which is a vector composed of multiple variables, expressed as: s(t) = [T(t), H(t), I(t), V(t)]. Wherein, T(t): the real-time temperature of the environment where the insect control device is located (temperature at time t) measured by the temperature sensor; H(t): the real-time humidity of the environment where the insect control device is located (humidity at time t) measured by the humidity sensor; I(t): the real-time light intensity at the location of the insect control device (light intensity or light irradiance at time t) measured by the light sensor; and V(t): the real-time wind speed of the environment where the insect control device is located (wind speed at time t) measured by the wind speed sensor. This environmental information may affect the insect control effect, and this environmental data is transmitted to the electronic control module 8 in real time for calculation of subsequent decision-making processes.

[0050] Step S102: Input the environmental state vector into the policy network to calculate the control parameters, which include 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.

[0051] Based on the current environmental state vector, the Actor network (i.e., the policy network) determines the next action, namely, adjusting the control parameters of the insect-killing device (i.e., adjusting 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 parameter a(t) to be adjusted includes the intensity P. light (t), temperature T heat (t) and band B(t), which represent the control action of the insecticidal device at time t. Specifically, the output of the Actor network is: a(t)=σ(MLP(s(t)))=[P light (t),T heat (t),B(t)]

[0052] Where MLP stands for fully connected; σ represents the sigmoid activation function, which restricts the output to a specific range (such as light intensity percentage and temperature range) to obtain the parameter to be controlled; P light (t) Intensity control (when the insect-attracting component is a high-pressure mercury lamp, and the insect-killing component is an iodine-tungsten lamp, this 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 needs; T heat(t) Temperature Control: When the insecticidal component is an iodine-tungsten lamp, the temperature of the iodine-tungsten lamp heat source is controlled to ensure the killing effect when humidity is high; B(t) Selecting the Appropriate Wavelength: When the insecticidal component is a high-pressure mercury lamp, the wavelength combination of the high-pressure mercury lamp light source is adjusted to maximize the pest capture rate and reduce the impact on non-target organisms. The electronic control module 8 comprehensively evaluates the collected environmental data through a neural network to generate the most suitable combination of control parameters for 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 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 wavelength (i.e., the high-pressure mercury lamp) to ensure that the insecticidal effect is not affected by changes in light and wind speed.

[0053] Step S103: Calculate the reward value by weighting the capture rate, energy consumption, and ecological impact index. The capture rate is the ratio of the number of target organisms captured per unit time to the total capture amount obtained from the analysis of images collected by the camera. The energy consumption is the energy consumption of the insecticidal device. The ecological impact index is the ratio of the number of non-target organisms captured to the total capture amount.

[0054] The electronic control module adjusts its operating status and executes corresponding actions based on the control parameters calculated by the strategy network. After execution, the system receives the following feedback. Taking into account indicators such as capture rate, energy consumption, and ecological impact index, a reward value is obtained through weighted calculation. The reward function for calculating the reward value is as follows: R(t) = α·capture rate(t) - β·energy consumption(t) - γ·ecological impact index(t)

[0055] The capture rate (t) represents the capture efficiency of the insecticidal device on target organisms at time t under the current environment. It is obtained by using a camera installed on the support component to monitor and record the number of captured pests (target organisms) and beneficial insects (non-target organisms) in real time, and then calculating the number of pests captured per unit time to arrive at the capture rate. Energy consumption (t) represents the energy consumption of the insecticidal device at time t. It is obtained by real-time monitoring of the current consumption of each power-consuming component of the device, such as the light source and heat source, using current detection, and combining this with voltage data to calculate the actual power consumption of each component, such as the light source and heat source, and summing the overall energy consumption. Ecological impact index (t) represents the impact of the insecticidal device on non-target organisms (such as beneficial insects) at time t. It is obtained by using a camera to monitor the number of non-target organisms (such as beneficial insects) and calculating the ecological impact index. In an exemplary embodiment, the unit time can be set to 10 minutes. Thus, every 10 minutes, the capture status is assessed, with 15 pests and 1 beneficial insect captured. The capture rate (t) is calculated as 15 / 16 × 100% = 93.75%, and the ecological impact index (t) is 1 / 16 × 100% = 6.25%. Furthermore, it should be noted that for the identification of pests and beneficial insects, in specific implementations, the open-source model YOLOv8 can be used to train on a dataset of pests and beneficial insects to derive a pest-benefit discriminator to identify the types of insects captured. This will not be elaborated further in this embodiment of the invention.

[0056] The weighting parameters α, β, and γ are used to balance capture efficiency, energy consumption, and ecological safety. These weighting parameters satisfy the following relationships: α + β + γ = 1, 0 < α, β, γ < 1, α ∈ [0.4, 0.6], β ∈ [0.3, 0.5], γ ∈ [0.1, 0.3].

[0057] α (capture efficiency weight) reflects the insecticidal device's ability to capture target pests; a higher α value indicates a greater focus on increasing the capture rate. β (energy consumption weight) reflects the energy consumption of insecticidal treatment; a higher β value indicates a greater focus on energy conservation and reducing energy consumption. γ (ecological impact weight) reflects the degree of impact of the insecticidal device on non-target organisms; a higher γ value indicates a greater focus on minimizing the impact on beneficial insects and other eco-friendly organisms. The overall goal of the insecticidal device is to maximize the reward function R(t), that is, to reduce energy consumption and minimize the impact on non-target organisms while ensuring efficient pest capture.

[0058] Step S104: Calculate the advantage function value based on the calculated reward value and the value estimate corresponding to the environmental state vector at different times, wherein the value estimate is a scalar value calculated by the evaluation network through a multi-layer neural network based on the input environmental state vector at different times.

[0059] In this embodiment of the invention, the evaluation network of the insect control device uses a Critic network 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 a nonlinear transformation through the hidden layers, and finally outputs a scalar value representing the value estimate V of the environment state vector. φ Similarly, after obtaining the new environment state vector s(t+1), i.e., 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 performance of the insect control device under these environmental conditions directly affects the final insect control efficiency, energy consumption, and impact on non-target organisms.

[0060] To further improve the performance of the pest control device, the Critic network calculates the dominance function by comparing the actual reward with the expected value, obtaining the dominance function value A(t), which guides the Actor network (i.e., the policy network) to optimize the control strategy. The formula for calculating the dominance function value A(t) is as follows: A(t) = R(t) + δ·V φ (s(t+1))-V φ (s(t))

[0061] Where R(t) is the reward value calculated at time t, including capture rate, energy consumption, and impact on non-target organisms, δ is a discount factor, 0≤δ<1, used to balance short-term and long-term rewards, and V φ (s(t+1)) and V φ (s(t)) represent the value estimates of the current and new environmental state vectors, respectively. In this way, the Critic network can estimate the "goodness" or "badness" of the current pest control device state. Specifically, the Critic network evaluates the "goodness" or "badness" of the current state and actions based on the sign and magnitude of the dominance function value A(t). A larger dominance function value indicates better performance of the current action in that state, and the Critic network considers the state "good." A smaller or negative dominance function value indicates poor performance of the current action in that state, and the Critic network considers the state "bad." The dominance function value is then fed back to the Actor network to improve the control strategy.

[0062] Step S105: The evaluation network inputs the calculated advantage function value into the policy network, and the policy network adjusts the control parameters according to the advantage function value.

[0063] In this embodiment of the invention, the Actor network is responsible for selecting appropriate control parameters (such as light source intensity, heat source temperature, and light source wavelength as described in the above embodiments) based on the current environmental state vector, and continuously optimizing its control strategy using the policy gradient method. Specifically, the Actor network determines how to adjust the various control parameters of the insecticidal device by estimating the current environmental state, thereby maximizing the reward function value R(t) and ensuring the device maintains optimal performance under environmental changes. The control parameter update formula of the Actor network is as follows:

[0064] Where, α actor This is the learning rate of the Actor network, which is configured to 0.0001 in this embodiment of the invention. It is the policy gradient, where the policy function π θ (a(t)|s(t)) represents the policy function π given the environment state vector s(t). θ Choose the probability distribution of the control parameter a(t). The policy parameter θ controls the shape and characteristics of this probability distribution. The strategy function indicates the probability of adjusting parameter θ to increase or decrease a specific control parameter, thereby optimizing the strategy. The strategy function represents how to update the control strategy to adapt to the current environment. A(t) is the dominance function value (i.e., the calculated result of the dominance function), used to measure the current controlled light source intensity P. light (t), heat source temperature T heat The insecticidal effect achieved by selecting the light source band B(t) and B(t) is as follows. With the update process of this learned control strategy, the Actor network can gradually achieve dynamic optimization, ensuring that the device performs optimally under various environmental conditions.

[0065] It should be noted that the steps described in the above embodiments are repeated iteratively, which can continuously optimize the control strategy and ensure that the insecticidal device maintains the best insecticidal effect and energy efficiency in a dynamic environment.

[0066] In the insect control method provided in this invention embodiment, the Critic network is responsible for evaluating the performance of the device under the current environmental conditions. Specifically, the Critic network evaluates the insect control device's effectiveness, energy efficiency, and impact on non-target organisms under specific operating conditions by calculating the "value" (i.e., value estimate) of the current environmental state. Based on this, the Critic network generates a feedback signal (i.e., the dominance function value) reflecting the performance of the current control strategy and provides 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, and light source band selection), thereby optimizing the control parameters of the insect control device, improving insect control efficiency, reducing energy consumption, and minimizing the impact on non-target organisms. In this way, the Critic network and the Actor network cooperate to form a closed-loop control system, ensuring that the insect control device can adjust its operating parameters in real time according to dynamic changes in the environment, thus always maintaining optimal operating conditions.

[0067] The coordinated adjustment of light source intensity, heat source temperature, and light source wavelength selection is optimized through the collaboration of Critic and Actor networks to ensure the coordinated operation of each parameter. For example, in environments with high temperature, high humidity, and low wind speed, the insecticidal device simultaneously increases the heat source temperature, maintains a high light source intensity, and selects a concentrated wavelength light source to ensure efficient pest control. In environments with low temperature, low humidity, and high wind speed, the device reduces the heat source temperature, lowers the light source intensity, and selects a broad wavelength light source to save energy while expanding the trapping range. Through this multi-light source combination and parameter synergistic optimization control strategy, the insecticidal device can achieve optimal pest control effect and energy efficiency balance under different environmental conditions.

[0068] Based on the description of the above embodiments, the insecticidal method provided by this invention, using the insecticidal device described in the above embodiments, achieves highly efficient insecticidal 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 utilizes a combination of high-pressure mercury lamps and tungsten iodine lamps, combined with an environmental sensing system and adaptive adjustment algorithm, to dynamically adjust the light source intensity, wavelength selection, and heat source temperature to adapt to environmental changes such as temperature, humidity, light intensity, and wind speed, ensuring that the insecticidal efficiency is always at its optimal state.

[0069] The intelligent control system of the insect control device integrates an environmental sensing sensor module, including temperature, humidity, light intensity, and wind speed sensors. This allows for real-time collection of external environmental conditions and transmission to an adaptive adjustment algorithm. The insect control method dynamically adjusts the light source intensity, wavelength combination, and heat source temperature of the high-pressure mercury lamp and tungsten iodine lamp based on the collected data. In practical use, for low-light environments, increasing the light source intensity of the high-pressure mercury lamp (e.g., from 50% to 100%) while selecting a wider wavelength (300-600 nm) can attract pests over a larger area, improving the capture rate. For high-light environments, reducing the light source intensity (e.g., from 100% to 50%) and selecting a more concentrated wavelength (350-450 nm) avoids energy waste while maintaining the trapping effect. In high-temperature and high-humidity environments, increasing the heat source temperature of the tungsten iodine lamp (e.g., from 500℃ to 750℃) needs to be coordinated with an appropriate light source intensity (e.g., maintaining a high light source intensity) to ensure that the high temperature can quickly kill pests without excessive energy consumption. In low-temperature and low-humidity environments, combining the reduction of heat source temperature (e.g., from 750℃ to 500℃) with the reduction of light source intensity (e.g., from 100% to 50%) can save energy while maintaining basic pest control effectiveness. Experiments show that environmentally-aware adaptive light source adjustment devices are significantly attractive to common agricultural pests (such as aphids, cotton bollworms, and fall armyworms), achieving a capture efficiency of over 95% and significantly improving pest control effects.

[0070] In summary, the embodiments of the present invention, by combining environmental sensing technology, multi-light source combination, and adaptive adjustment algorithm, realize an efficient, intelligent, and environmentally friendly pest control device and a pest control method based on the device. This not only maintains excellent pest control effects in various complex environments, but also significantly reduces energy consumption and minimizes adverse impacts on the ecological environment, providing an innovative solution for pest control in agriculture, forestry, aquaculture, and other fields.

[0071] Through the description of the above method embodiments, those skilled in the art can clearly understand that the present invention can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0072] This invention provides a non-volatile computer storage medium storing computer-executable instructions that can execute the insect control method in any of the above-described method embodiments.

[0073] The above-described insect control method is based on the insect control device provided in the embodiments of the present invention, and possesses the corresponding functional modules and beneficial effects of the insect control device. Technical details not described in detail in this embodiment can be found in the insect control device provided in the embodiments of the present invention.

[0074] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.

[0075] The various embodiments in this specification are described in a progressive manner. Similar or identical parts between embodiments can be referred to mutually. Each embodiment focuses on describing the differences from other embodiments. In particular, for apparatus or system embodiments, since they are basically similar to method embodiments, the description is relatively simple; relevant parts can be referred to the descriptions in the method embodiments. The apparatus and system embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without creative effort.

[0076] It should be noted that, in this document, relational terms such as "first" and "second" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0077] The above description is merely a specific embodiment of the present invention, enabling those skilled in the art to understand or implement the invention. Various modifications to these embodiments will be readily 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 invention. Therefore, the present invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. An insect eradication device, characterized in that, The device comprises a base, a support assembly, a shell, a collection plate, an electronic control module and a sensor module. The base is movably connected with the shell through the support assembly. The shell is provided with a moth-attracting assembly and a moth-eliminating assembly, and comprises a concave cavity for forming a high-temperature cavity during operation. The collection plate is arranged on the support assembly and used for collecting insects. The sensor module comprises a temperature and humidity sensor, an illumination sensor, a wind speed sensor and a camera, and the temperature and humidity sensor, the illumination sensor, the wind speed sensor and the camera are in communication connection with the electronic control module.

2. The device of claim 1, wherein, The temperature and humidity sensor is used for collecting the temperature and humidity of the environment, the illumination sensor is used for collecting the light intensity of the environment, the wind speed sensor is used for collecting the wind speed of the environment, and the camera is used for collecting the image of the collection plate.

3. The device of claim 1, wherein, The electronic control module is in communication connection with the moth-attracting assembly and the moth-eliminating assembly, and the control parameters of the moth-attracting assembly and the moth-eliminating assembly are calculated and adjusted through a reinforcement learning algorithm according to the collected temperature, humidity, light intensity, wind speed and image.

4. The device of claim 1, wherein the device is configured to emit a sound at a frequency of 20- 2000 Hz. The temperature and humidity sensor and the illumination 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 collection plate on the support assembly and used for collecting the image on the collection plate.

5. A method of killing insects based on the device according to claims 1-4, characterized in that, The moth-attracting assembly comprises a high-pressure mercury lamp arranged at the center of the inner wall of the concave cavity, and the moth-eliminating assembly comprises an iodine-tungsten lamp arranged at the edge of the inner wall of the concave cavity. The support assembly comprises a support rod and a support frame, one end of the support rod is arranged on the top surface of the base, the other end of the support rod is connected with the support frame, and the support frame is movably connected with the shell through a rotating shaft so that the shell can rotate at an angle. The device comprises: The temperature, humidity, light intensity and wind speed are obtained through the temperature and humidity sensor, the illumination sensor and the wind speed sensor, and an environment state vector is formed; The environment state vector is input into a policy network to calculate control parameters, the control parameters comprising the intensity of the moth-attracting assembly and the moth-eliminating assembly, the temperature of the moth-eliminating assembly and the wave band of the moth-attracting assembly; The reward value is calculated by weighting the capture rate, the energy consumption and the 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 image collected by the camera, the energy consumption is the energy consumption of the moth-eliminating device, and the ecological impact index is the ratio of the number of non-target organisms captured to the total capture amount; 6. The insect control method according to claim 5, characterized in that, The environmental state vector is input into the policy network, and a control parameter is calculated, as follows: a(t) = σ(MLP(s(t))) = [P light (t), T heat (t), B(t)] Wherein, a(t) represents a control parameter, sigma represents a Sigmoid activation function, MLP represents full connection, s(t) represents an environment state vector at t moment, P light (t) represents a calculated intensity at t moment, T heat (t) represents a temperature at t moment, B(t) represents a wave band at t moment, and the policy network comprises an Actor network.

7. The method of claim 5, wherein, The advantage function value is calculated 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 a 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 policy network, and the policy network adjusts the control parameters according to the advantage function value. The reward value is calculated by weighting the capture rate, the energy consumption and the ecological impact index according to the following formula: R(t) = a·capture rate(t) - β·energy consumption(t) - γ·ecological impact index(t) Wherein, R(t) represents the reward value at time t, the capture rate(t) represents the capture rate at time t, the energy consumption(t) represents the energy consumption at time t, the ecological influence index(t) represents the ecological influence index at time t, alpha represents the capture rate weight, beta represents the energy consumption weight, gamma represents the ecological influence weight.

8. The method of claim 7, wherein, The capture rate weight, the energy consumption weight and the ecological influence weight are valued as follows: alpha+beta+gamma=1 0<alpha, beta, gamma<1 alpha belongs to [0.4, 0.6], beta belongs to [0.3, 0.5], gamma belongs to [0.1, 0.3].

9. The method of claim 5, wherein, The advantage function value is calculated according to the calculated reward value and the value estimation corresponding to the environment state vector at different time, and the formula is as follows: A(t) = R(t) + δ·V φ (s(t+1))-V φ (s(t)) wherein A(t) represents a value of the advantage function at time t, R(t) represents a reward value at time t, V φ (s(t+1)) represents a value estimate at time t+1, s(t+1) represents an environment state vector at time t+1, V φ (s(t)) represents a value estimate at time t, s(t) represents an environment state vector at time t, δ is a discount factor, and 0≤δ<1, the evaluation network comprising a Critic network.

10. The method of claim 5, wherein, The evaluation network inputs the calculated advantage function value into the policy network, and the policy network adjusts the control parameter according to the advantage function value, as follows: wherein θ t+1 denotes the policy parameter at time t+1, used to control the shape and characteristics of the selection of the control parameter a(t) at time t+1; θ t denotes the policy parameter at time t; a actor denotes the learning rate of the policy network; representing a policy gradient, denotes the probability of adjusting the policy parameter, π θ (a(t)|s(t)) denotes the policy function π θ The probability distribution over control parameters a(t) is chosen, A(t) is the advantage function value.

Images