Hotel pest killing robot capable of autonomous charging and chemical liquid management

Abstract

The present application relates to the technical field of intelligent disinfection and killing equipment, and particularly relates to a hotel pest killing robot capable of autonomous charging and medicine management, which comprises an autonomous charging and medicine management system, a sensor cluster, an atomizing nozzle, a charging port, a display screen, a driver, a mobile chassis, a balancer and a medicine tank, the autonomous charging and medicine management system comprises an environment perception module, a pest prediction module, an energy management module, a medicine management module and a load management module, the system carries out feature extraction and model identification according to pest environment monitoring data to obtain a pest density prediction distribution, and carries out charging and liquid supplementing decision according to task energy consumption estimation and medicine residual amount entropy, and carries out joint scheduling optimization according to charging urgency and liquid supplementing urgency, so as to realize energy and medicine collaborative management and multi-robot charging load balancing, thereby improving the intelligent level of hotel pest killing and the overall efficiency of the system.

Description

Technical Field

[0001] This invention relates to the field of intelligent pest control equipment technology, and in particular to a hotel pest control robot that can autonomously charge and manage pesticide solutions. Background Technology

[0002] While mobile spraying equipment is currently used in hotel pest control, it still has shortcomings. First, existing pest control devices mostly rely on manual battery replacement and manual pesticide replenishment, lacking an autonomous energy and pesticide management mechanism designed for the characteristics of hotel scenarios: "long operation time, wide coverage area, and rapid energy and pesticide consumption." This leads to frequent manual maintenance and low equipment utilization. Second, there is a lack of intelligent prediction capabilities for pest density distribution, making it impossible to predict high-risk areas in advance based on environmental data. This results in unreasonable allocation of pest control resources and delayed control effects. Third, charging and replenishment decisions are independent, without considering their synergistic relationship. When multiple robots need to serve simultaneously, it can easily cause base station port congestion and excessively long service waiting times. In addition, the lack of a quantitative assessment system for charging and replenishment urgency makes it impossible to achieve load balancing in multi-robot scenarios, causing some robots to wait excessively and miss the optimal pest control opportunity.

[0003] Chinese patent application CN119429110A discloses a drone-based pesticide spraying device and management system. The device and system include a spraying system and a control system. The spraying system comprises a liquid storage system, a pesticide delivery system, and a pesticide spraying system. The liquid storage system is connected to the pesticide spraying system via the pesticide delivery system, and the control system is connected to both the pesticide delivery system and the pesticide spraying system. The liquid storage system includes a pesticide tank mounted on the drone body. The pesticide delivery system includes components connected to the pesticide tank. The cover plate is connected to the tank, and a pump body is installed inside the tank. A connecting pipe is installed on the pump body. The pesticide spraying system includes a nozzle installed on the drone shell, and the nozzle is connected to the connecting pipe. The control system includes a stirring structure, a power cord 1, a power cord 2, and a controller installed inside the tank, which makes it easy to fix tanks of different sizes and to stir the pesticides in the tank. It can be seen that this solution still has the problems of insufficient pest control efficiency in hotels due to the lack of pest density prediction, energy and pesticide liquid collaborative management, and multi-robot charging load balancing scheduling. Summary of the Invention

[0004] Therefore, this invention provides a hotel pest control robot that can autonomously charge and manage pesticide solutions, in order to overcome the problems of insufficient pest control efficiency in hotels caused by the lack of pest density prediction, energy and pesticide collaborative management, and balanced scheduling of multi-robot charging load in the prior art.

[0005] To achieve the above objectives, the present invention provides a hotel pest control robot capable of autonomous charging and pesticide management, comprising:

[0006] The autonomous charging and pesticide management system, connected to the mobile chassis, is used to acquire pest environmental monitoring data and task status data, acquire pest density prediction distribution, acquire expected task energy consumption and pesticide residue entropy, and output charging trigger decisions and replenishment trigger decisions based on expected task energy consumption and pesticide residue entropy, acquire charging urgency and replenishment urgency, acquire pesticide-charging urgency score and port comprehensive load index, and output to the target base station according to the principle of minimizing maximum load;

[0007] A sensor cluster, connected to a mobile chassis and an autonomous charging and pesticide management system, is used to collect pest environmental monitoring data.

[0008] Atomizing nozzle, which is connected to the medicine tank, is used to spray the medicine.

[0009] The charging port connects to the mobile chassis and is used to connect to the charging port for charging.

[0010] The display screen, connected to the balancer, is used to display charging trigger decisions and replenishment trigger decisions;

[0011] A drive unit, connected to the mobile chassis, is used to drive the mobile chassis to move based on charging trigger decisions and replenishment trigger decisions;

[0012] The mobile chassis, which is connected to the autonomous charging and medicine management system, sensor cluster, charging port, driver, balancer and medicine tank, is used to carry the device for movement;

[0013] A balancer, which connects to the mobile chassis and display screen, is used to maintain balance;

[0014] The medicine tank, which is connected to the mobile chassis and atomizing nozzle, is used to store the medicine.

[0015] Furthermore, the autonomous charging and medication management system includes:

[0016] The environmental sensing module is used to acquire pest environmental monitoring data and task status data;

[0017] The pest prediction module is used to extract features from pest environmental monitoring data to obtain pest environmental feature vectors. It is also used to perform model recognition based on pest environmental feature vectors to obtain the predicted distribution of pest density.

[0018] The energy management module is used to acquire task status data, estimate task energy consumption based on the task status data, obtain the expected energy consumption of the task, acquire the remaining power and the safe return power threshold, and output the charging trigger decision based on the expected energy consumption, remaining power and safe return power threshold.

[0019] The pesticide management module is used to obtain the pesticide residue entropy based on the predicted distribution of pest density, and output the replenishment trigger decision based on the pesticide residue entropy.

[0020] The load management module is used to obtain the charging urgency based on the safe return-to-home power threshold, the replenishment urgency based on the liquid level entropy, and to obtain a liquid-charging urgency score based on both the charging urgency and replenishment urgency. It is also used to obtain the port comprehensive load index and to optimize the acquisition process of the port comprehensive load index based on the charging trigger decision and the replenishment trigger decision. Finally, it outputs the target base station based on the liquid-charging urgency score and the port comprehensive load index by minimizing the maximum load principle.

[0021] Furthermore, the pest prediction module extracts features from the pest environmental monitoring data, specifically by: normalizing the temperature and humidity data to obtain normalized ambient temperature T' and normalized relative humidity H'; performing a logarithmic transformation on the illuminance data L to obtain transformed illuminance data L', setting L'=ln(L+1); standardizing the carbon dioxide concentration data C to obtain standardized carbon dioxide concentration data C'; encoding and converting the human presence data P to obtain encoded human presence data P'; and concatenating the normalized ambient temperature T', normalized relative humidity H', transformed illuminance data L', standardized carbon dioxide concentration data C', and encoded human presence data P' to obtain the pest environmental feature vector E={T',H',L',C',P'}.

[0022] Furthermore, the pest prediction module performs model recognition based on the pest environmental feature vector. Specifically, it constructs a pest density prediction model by inputting the pest environmental feature vector E into the pest density prediction model to obtain the predicted pest density distribution output by the pest density prediction model. ={ , ,..., },in, Indicates the order of regions. =1,2,……, , The total number of grid units for the hotel area.

[0023] Furthermore, the energy management module acquires task status data and estimates task energy consumption based on the task status data. Specifically, the task status data is mapped to the [0,1] interval using the maximum-minimum normalization method to obtain the standard value Sq of the area to be eliminated, the standard value Davg of the predicted pest density, the standard value Ctm of the terrain complexity, and the standard value Uem of the task urgency.

[0024] Based on the standard value Sq of the area to be exterminated, the standard value Davg of the predicted pest density, the standard value Ctm of the terrain complexity, the standard value Uem of the task urgency, and the energy consumption weight set We={we1,we2,we3,we4}, the expected energy consumption F1 of the task is calculated, and F1 is set as (Sq×Davg×we1+Ct×we2+Ue×we3)×we4, and we1+we2+we3=1.

[0025] Furthermore, the energy management module acquires the remaining battery power and the safe return battery power threshold, specifically by: acquiring the remaining battery power F2 in real time through the battery management system, and determining the current location coordinates of the device (…). , ), the coordinates of the nearest base station ( , Energy consumption per unit distance for return and return safety factor The safe return-to-home battery threshold F3 is calculated and set as F3= , and 1 < <2;

[0026] The energy management module outputs a charging trigger decision based on the expected energy consumption F1, remaining power F2, and safe return power threshold F3, wherein:

[0027] When F1+F3<F2×η1, no output is made for the charging trigger decision;

[0028] When F1+F3≥F2×η1, the charging trigger decision is output to the driver, where η1 is the power safety margin coefficient and 0<η1<1, and the charging trigger decision is pushed to the display screen.

[0029] Furthermore, the pesticide management module obtains the entropy of the remaining pesticide solution based on the predicted distribution of pest density. Specifically, it calculates the entropy of the remaining pesticide solution in the region. percentage To obtain, among which, Indicates the order of regions and predicts distribution based on pest density. and the proportion of the remaining drug solution in region i The entropy Y of the remaining liquid medicine is calculated, and Y is set as follows: ;

[0030] The remaining entropy of the medicine solution Y is compared with the preset remaining entropy of the medicine solution Y0. Based on the comparison result, the state of the remaining entropy of the medicine solution is judged, and the replenishment trigger decision is output based on the judgment result. Among them:

[0031] When Y≤Y0, the liquid management module determines that the state of the liquid remaining entropy is acceptable and does not output the liquid replenishment trigger decision;

[0032] When Y > Y0, the liquid management module determines that the state of the liquid remaining entropy is unacceptable, outputs the liquid replenishment trigger decision to the driver, and pushes the liquid replenishment trigger decision to the display screen.

[0033] Furthermore, the load management module obtains the charging urgency based on the safe return-to-home power threshold, specifically by calculating the charging urgency Uc based on the remaining power F2, the safe return-to-home power threshold F3, and the charging urgency weighting coefficient wc, and setting Uc=wc×(F3 / F2).

[0034] The load management module obtains the urgency of replenishment based on the drug liquid remaining entropy, and calculates the urgency of replenishment Uf based on the drug liquid remaining entropy Y, the preset drug liquid remaining entropy Y0 and the replenishment urgency weighting coefficient wf, and sets Uf=wf×(Y / Y0).

[0035] The load management module outputs a drug solution-charging urgency score based on the charging urgency and the replenishment urgency. Specifically, it calculates the drug solution-charging urgency score S based on the charging urgency Uc, the replenishment urgency Uf, the charging demand identifier Dc, and the replenishment demand identifier Df, and sets S = Dc × Uc + Df × Uf.

[0036] Furthermore, the load management module obtains the comprehensive load index of the ports, specifically by obtaining the candidate base station set Zi={Z1,Z2,...,Zn}, where n represents the order, n=1,2,...,n, and obtaining the charging port occupancy rate Ri, liquid replenishment port occupancy rate Hi, expected waiting time Ti, and distance Di of the candidate base station set Zi.

[0037] The port comprehensive load index Li is calculated based on the charging urgency Uc, the liquid replenishment urgency Uf, the charging port occupancy rate Ri, the liquid replenishment port occupancy rate Hi, the expected waiting time Ti, the distance Di, and the load weight set Wl={wl1,wl2,wl3,wl4}. The formula is: Li=wl1×(Ri×Uc+Li×Uf)+wl2×Ti+wl3×Di+wl4×(Uc+Uf);

[0038] The load management module optimizes the process of obtaining the port comprehensive load index based on the charging trigger decision and the liquid replenishment trigger decision, wherein:

[0039] When the output charging trigger decision is not outputting the liquid replenishment trigger decision, the process of obtaining the port comprehensive load index is optimized by weighting: the liquid replenishment urgency weight is reset to 0, Uf'=0 is set, and the port comprehensive load index Li is recalculated.

[0040] When no charging trigger decision is output but a liquid replenishment trigger decision is output, the process of obtaining the port comprehensive load index is optimized by weighting: the charging urgency weight is reset to 0, UC'=0 is set, and the port comprehensive load index Li is recalculated.

[0041] When the output charging trigger decision and the output liquid replenishment trigger decision are both made, the process of obtaining the port comprehensive load index is not optimized by weight.

[0042] Furthermore, the load management module outputs the target base station based on the liquid-charging urgency score and the port comprehensive load index according to the principle of minimizing the maximum load. The base station that minimizes max(Li) is selected as the target base station according to the port comprehensive load index Li. When there are multiple base stations with the same max(Li), the base station with the shortest distance Di is selected as the target base station.

[0043] Compared with the prior art, the beneficial effects of the present invention are as follows: the robot is supported by a mobile chassis and achieves planar movement; the driver achieves precise drive control; the charging port connects to the charging port for energy replenishment; the liquid tank stores the liquid; the atomizing nozzle performs the spraying of the agent; the sensor cluster comprehensively perceives environmental information; the display screen realizes human-computer interaction; and the autonomous charging and liquid management system integrates intelligent algorithms to achieve pest density prediction, energy and liquid coordinated management, and charging load balancing, so as to build a complete intelligent pest control hardware platform and adapt to the high mobility operation requirements of the complex environment of the hotel.

[0044] In particular, the system uses an environmental perception module to collect pest environmental monitoring data and task status data from multiple sources, providing a comprehensive and accurate data foundation for subsequent pest prediction and energy / pesticide management, thereby improving the accuracy of pest control decisions. The system also uses a pest prediction module to obtain pest density predictions and distributions through environmental feature extraction and model recognition, enabling early identification of high-risk areas and rational allocation of pest control resources, thus improving the timeliness and accuracy of control effects. Furthermore, the system uses an energy management module to make intelligent charging decisions based on task energy consumption estimates and power safety thresholds. The system also uses a pesticide management module to trigger replenishment based on pesticide residue entropy, enabling collaborative management and autonomous maintenance of energy and pesticides, thereby improving equipment utilization and operational continuity. Finally, the system uses a load management module to perform joint scoring based on charging urgency and replenishment urgency, and performs weight optimization according to the trigger decision type, selecting target base stations by minimizing the maximum load principle, thereby achieving charging load balancing and global scheduling optimization in multi-robot scenarios, thus improving the system's pest control efficiency and service response speed in multi-robot hotel scenarios. Attached Figure Description

[0045] Figure 1 This is a schematic diagram of the structure of the hotel pest control robot that can autonomously charge and manage pesticide solutions in this embodiment;

[0046] Figure 2 This is a schematic diagram of the self-charging and liquid medicine management system of this embodiment;

[0047] Figure 3 This is a flowchart illustrating the task energy consumption prediction method in this embodiment. Detailed Implementation

[0048] To make the objectives and advantages of the present invention clearer, the present invention will be further described below with reference to embodiments; it should be understood that the specific embodiments described herein are merely for explaining the present invention and are not intended to limit the present invention.

[0049] Preferred embodiments of the present invention will now be described with reference to the accompanying drawings. Those skilled in the art should understand that these embodiments are merely illustrative of the technical principles of the present invention and are not intended to limit the scope of protection of the present invention.

[0050] It should be noted that in the description of this invention, the terms "upper", "lower", "left", "right", "inner", "outer", etc., which indicate directions or positional relationships, are based on the directions or positional relationships shown in the accompanying drawings. This is only for the convenience of description and is not intended to indicate or imply that the device or element must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, it should not be construed as a limitation of this invention.

[0051] Furthermore, it should be noted that, in the description of this invention, unless otherwise explicitly specified and limited, the terms "installation," "connection," and "linking" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection of two components. Those skilled in the art can understand the specific meaning of the above terms in this invention according to the specific circumstances.

[0052] Please see Figure 1 As shown, this is a structural diagram of the hotel pest control robot capable of autonomous charging and pesticide management in this embodiment, including:

[0053] The autonomous charging and pesticide management system 1, connected to the mobile chassis 7, is used to acquire pest environmental monitoring data and task status data, acquire pest density prediction distribution, acquire expected task energy consumption and pesticide residue entropy, and output charging trigger decision and replenishment trigger decision based on expected task energy consumption and pesticide residue entropy, acquire charging urgency and replenishment urgency, acquire pesticide-charging urgency score and port comprehensive load index, and output to the target base station through the principle of minimizing maximum load;

[0054] Sensor cluster 2, which is connected to mobile chassis 7 and autonomous charging and pesticide management system 1, is used to collect pest environmental monitoring data;

[0055] Atomizing nozzle 3 is connected to the liquid tank 9 and is used to spray liquid medicine;

[0056] Charging port 4 is connected to the mobile chassis 7 and is used to connect to the charging port for charging.

[0057] Display screen 5, which is connected to balancer 8, is used to display charging trigger decision and liquid replenishment trigger decision;

[0058] A drive 6, which is connected to a mobile chassis 7, is used to drive the mobile chassis 7 to move according to charging trigger decisions and replenishment trigger decisions;

[0059] The mobile chassis 7 is connected to the autonomous charging and medicine management system 1, the sensor cluster 2, the charging port 4, the driver 6, the balancer 8, and the medicine tank 9, and is used to carry the device for movement.

[0060] A balancer 8, which is connected to the movable chassis 7 and the display screen 5, is used to maintain balance;

[0061] The medicine tank 9, which is connected to the mobile chassis 7 and the atomizing nozzle 3, is used to store the medicine.

[0062] Specifically, the sensor cluster includes a temperature and humidity sensor, a light intensity sensor, a carbon dioxide concentration sensor, and a human presence infrared sensor. The device refers to a hotel pest control robot that can autonomously charge and manage pesticide solutions. The sensor cluster is connected to the autonomous charging and pesticide solution management system via internal connection lines. This embodiment does not limit the type of internal connection lines, and those skilled in the art can freely choose according to actual needs, such as a CAN bus.

[0063] Specifically, the self-charging and pesticide management-enabled hotel pest control robot can be applied to pest control scenarios in multi-robot hotel environments, such as guest rooms, restaurants, kitchens, and corridors. The robot is supported by a mobile chassis for planar movement, with a driver providing precise drive control, a charging port for energy replenishment, a pesticide tank for storing pesticides, a misting nozzle for spraying pesticides, a sensor cluster for comprehensive environmental information perception, and a display screen for human-computer interaction. The self-charging and pesticide management system integrates intelligent algorithms to predict pest density, manage energy and pesticides collaboratively, and balance charging load, thus facilitating the construction of a complete intelligent pest control hardware platform to meet the high mobility requirements of complex hotel environments.

[0064] Please see Figure 2 As shown, this is a schematic diagram of the autonomous charging and medication management system of this embodiment. The autonomous charging and medication management system includes:

[0065] The environmental sensing module is used to acquire pest environmental monitoring data and task status data;

[0066] The pest prediction module is used to extract features from pest environmental monitoring data to obtain pest environmental feature vectors. It is also used to perform model recognition based on pest environmental feature vectors to obtain the predicted distribution of pest density.

[0067] The energy management module is used to acquire task status data, and to estimate task energy consumption based on the task status data using a task energy consumption prediction method to obtain the expected task energy consumption. It is also used to acquire the remaining power and the safe return power threshold, and to output the charging trigger decision based on the expected task energy consumption, the remaining power, and the safe return power threshold.

[0068] The pesticide management module is used to obtain the pesticide residue entropy based on the predicted distribution of pest density, and output the replenishment trigger decision based on the pesticide residue entropy.

[0069] The load management module is used to obtain the charging urgency based on the safe return-to-home power threshold, the replenishment urgency based on the liquid level entropy, and to obtain a liquid-charging urgency score based on both the charging urgency and replenishment urgency. It is also used to obtain the port comprehensive load index and to optimize the acquisition process of the port comprehensive load index based on the charging trigger decision and the replenishment trigger decision. Finally, it outputs the target base station based on the liquid-charging urgency score and the port comprehensive load index by minimizing the maximum load principle.

[0070] Specifically, the autonomous charging and pesticide management system can be applied to hotel pest control robots that manage autonomous charging and pesticide application. The system extracts features and identifies models based on pest environmental monitoring data to predict pest density distribution. It then makes charging and replenishment decisions based on task energy consumption estimates and pesticide residue entropy. Furthermore, it performs joint scheduling optimization based on charging and replenishment urgency to achieve coordinated energy and pesticide management and balanced charging load among multiple robots. This improves the intelligence level and overall efficiency of hotel pest control. The system uses an environmental perception module to collect pest environmental monitoring data from multiple sources, providing a comprehensive and accurate data foundation for subsequent pest prediction and energy / pesticide management, thereby improving the accuracy of pest control decisions. The system also... The pest prediction module uses environmental feature extraction and model recognition to predict the distribution of pest density, enabling early identification of high-risk areas and rational allocation of pest control resources, thereby improving the timeliness and accuracy of pest control. The system also utilizes an energy management module to make intelligent charging decisions based on task energy consumption estimates and power safety thresholds. Furthermore, a pesticide management module uses pesticide residue entropy to trigger replenishment, facilitating collaborative management and autonomous maintenance of energy and pesticides, thus improving equipment utilization and operational continuity. Finally, a load management module performs joint scoring based on charging urgency and replenishment urgency, and optimizes weights according to trigger decision types. It selects target base stations by minimizing the maximum load principle, enabling balanced charging load and global scheduling optimization in multi-robot scenarios, thereby improving pest control efficiency and service response speed in multi-robot hotel environments.

[0071] Specifically, the environmental sensing module acquires pest environmental monitoring data and task status data.

[0072] Specifically, the pest environmental monitoring data includes temperature and humidity data, light intensity data, carbon dioxide concentration data, and human presence data. The temperature and humidity data refers to the ambient temperature and relative humidity collected by a temperature and humidity sensor, expressed in degrees Celsius and percentage. The light intensity data refers to the ambient light intensity collected by a light intensity sensor, expressed in lux. The carbon dioxide concentration data refers to the ambient carbon dioxide concentration collected by a carbon dioxide concentration sensor, expressed in parts per million (ppm). The human presence data refers to the activity status of personnel collected by a human presence infrared sensor, including both present and absent states. The task status data includes the area to be eradicated, the predicted average pest density, the terrain complexity coefficient, and the task urgency level. The area to be eradicated refers to the planar area that the current task needs to cover, expressed in square meters. In this embodiment, the area to be eradicated is obtained by accumulating data from a hotel area grid. The area, the predicted average pest density refers to the average value of the predicted pest density in the area to be eradicated. In this embodiment, the predicted average pest density is obtained by the arithmetic mean of the predicted pest density values ​​of the grid units to be eradicated in the hotel area. The terrain complexity coefficient is a coefficient that measures the terrain complexity of the area to be eradicated. The larger the value, the more complex the terrain. In this embodiment, the specific value of the terrain complexity coefficient Ct is not limited. Those skilled in the art can freely choose according to actual needs, such as by evaluating the area obstacle density and slope information, setting flat corridor Ct=1.0, complex guest rooms Ct=1.3, and narrow kitchen Ct=1.5. The task urgency level Ue refers to the task urgency level pre-assessed based on time sensitivity and hygiene requirements. The larger the value, the more urgent the task. In this embodiment, the routine prevention task Ue=1, the daily maintenance task Ue=2, and the emergency task Ue=3 are set.

[0073] Specifically, the environmental perception module collects pest environmental monitoring data through multi-source sensor fusion, so as to provide a comprehensive and accurate data foundation for subsequent pest prediction and energy and pesticide management, thereby improving the accuracy of pest control decisions.

[0074] Specifically, the pest prediction module extracts features from pest environmental monitoring data, specifically by: normalizing temperature and humidity data to obtain normalized ambient temperature T' and normalized relative humidity H'; performing a logarithmic transformation on illuminance data L to obtain transformed illuminance data L', setting L'=ln(L+1); standardizing carbon dioxide concentration data C to obtain standardized carbon dioxide concentration data C'; encoding and converting human presence data P to obtain encoded human presence data P'; and concatenating the normalized ambient temperature T', normalized relative humidity H', transformed illuminance data L', standardized carbon dioxide concentration data C', and encoded human presence data P' to obtain the pest environmental feature vector E={T',H',L',C',P'}.

[0075] Specifically, the normalization process refers to the process of unifying environmental data of different dimensions into the same numerical range. In this embodiment, the Min-Max normalization algorithm is used for normalization. The logarithmic transformation refers to the process of nonlinearly compressing illuminance data. In this embodiment, the natural logarithm function is used for logarithmic transformation. The standardization process refers to the process of converting carbon dioxide concentration data into a standard normal distribution. In this embodiment, the Z-score standardization algorithm is used for standardization. The encoding conversion refers to the process of converting the human presence state into a numerical encoding, such as setting P'=1 when there is someone and setting P'=0 when there is no one. The pest environmental feature vector refers to a numerical vector that comprehensively represents the environmental state.

[0076] Specifically, the pest prediction module constructs a pest environmental feature vector by extracting multi-dimensional environmental features, so as to fully characterize the environmental factors affecting pest activity and thus improve the accuracy of pest density prediction.

[0077] Specifically, the pest prediction module performs model recognition based on the pest environmental feature vector. Specifically, it constructs a pest density prediction model by inputting the pest environmental feature vector E into the pest density prediction model to obtain the predicted pest density distribution output by the model. ={ , ,..., },in, Indicates the order of regions. =1,2,……, , The total number of grid units for the hotel area.

[0078] Specifically, the pest density prediction model refers to a deep learning regression model that uses pest environmental feature vectors as input data and pest density prediction distribution as output data. The process of constructing the pest density prediction model in this embodiment includes: initializing the fully connected neural network by setting the input layer dimension to 5, the number of neurons in the first hidden layer to 128 with the ReLU activation function, the number of neurons in the second hidden layer to 64 with the ReLU activation function, the number of neurons in the third hidden layer to 32 with the ReLU activation function, the number of neurons in the output layer to m with the Sigmoid activation function, and setting the Dropout rate to 0.3. The historical environmental feature dataset is then divided into a 70% training set, a 20% validation set, and a 10% test set. Adam is selected as the optimizer, with a learning rate of 0.001, a batch size of 32, and 200 training epochs. The training set is input into the fully connected neural network for forward propagation, the mean squared error loss is calculated, and backpropagation and parameter updates are performed based on the loss to obtain the updated fully connected neural network. The validation set is then used for... The fully connected neural network is updated using the input set. The fully connected neural network with the smallest root mean square error on the validation set is used as the validated fully connected neural network. The validated fully connected neural network is then used as the input for the test set to evaluate the model. The validated fully connected neural network that passes the evaluation is output as the pest density prediction model. The historical environmental feature dataset refers to the set of historically collected pest environmental feature vectors and their corresponding actual pest density distributions. In this embodiment, the pest environmental feature vectors and their corresponding actual pest density distributions occurring in the past 6 months are used as the historical environmental feature dataset. The actual pest density distribution is obtained through manual inspection and counting with sticky insect boards. Passing the evaluation means that the root mean square error of the test set is ≤0.5 pests / square meter and the coefficient of determination R² is ≥0.85. The predicted pest density distribution refers to the predicted spatial distribution map of pest density in various areas of the hotel obtained through the pest density prediction model. This embodiment does not limit the specific method of dividing the hotel area into grids; those skilled in the art can freely choose according to actual needs, such as dividing the hotel floor into 10m × 10m grid units. This represents the pest density value in the first grid of the hotel area. This represents the pest density value in the second grid of the hotel area. Indicates the hotel area number The pest density value in each grid.

[0079] Specifically, the pest prediction module predicts pest density distribution by constructing a deep learning regression model, so as to identify high-risk areas in advance to predict energy consumption and pesticide dosage, thereby improving the timeliness and spatial accuracy of prevention and control effects.

[0080] Specifically, the energy management module acquires task status data and uses a task energy consumption prediction method to predict task energy consumption based on the task status data, thereby obtaining the expected energy consumption of the task.

[0081] Specifically, the energy management module acquires the remaining battery power and the safe return battery power threshold, specifically by: acquiring the remaining battery power F2 in real time through the battery management system, and determining the current location coordinates of the device ( , ), the coordinates of the nearest base station ( , Energy consumption per unit distance for return and return safety factor The safe return-to-home battery threshold F3 is calculated and set as F3= , and 1 < <2;

[0082] The energy management module outputs a charging trigger decision based on the expected energy consumption F1, remaining power F2, and safe return power threshold F3, wherein:

[0083] When F1+F3<F2×η1, no output is made for the charging trigger decision;

[0084] When F1+F3≥F2×η1, the charging trigger decision is output to the driver, where η1 is the power safety margin coefficient and 0<η1<1, and the charging trigger decision is pushed to the display screen.

[0085] Specifically, the remaining battery power refers to the current available battery power, measured in kilowatt-hours. In this embodiment, the remaining battery power F2 is obtained through the battery management system. The current device location coordinates refer to the device's location coordinates in the hotel coordinate system at the current moment. In this embodiment, the current device location coordinates are obtained through GPS. This indicates the value of the device's current position coordinates on the X-axis of the hotel coordinate system. This represents the value of the device's current location coordinates on the Y-axis of the hotel coordinate system. The nearest base station location coordinates refer to the location coordinates of the base station closest to the device's current location in the hotel coordinate system. In this embodiment, the nearest base station is obtained by querying a preset base station location database and calculating the Euclidean distance. This represents the coordinates of the nearest base station on the X-axis of the hotel's coordinate system. This represents the coordinates of the nearest base station on the Y-axis of the hotel coordinate system. This embodiment does not limit the specific construction method of the hotel coordinate system; those skilled in the art can freely choose according to actual needs, such as using the hotel reception location as the origin and the due east direction as the X-axis to establish the hotel coordinate system.

[0086] The unit distance return energy consumption refers to the electricity consumed by the device per unit distance traveled, measured in kilowatt-hours per meter. This embodiment obtains this energy consumption data through historical return energy consumption statistics. =0.005kWh / m, where the return safety factor refers to the safety multiple of electricity reserved to cope with emergencies. In this embodiment, the return process under different wind speeds and loads was simulated in a controlled environment. It was found that under the most common severe operating conditions, the actual energy consumption is about 30% higher than the theoretical energy consumption. Therefore, the following setting is made: =1.3, the power safety margin coefficient refers to the proportion of power reserved to extend battery life and cope with estimation errors. In this embodiment, η1 is set to 0.85 based on the fact that the old battery can release 85% of its nominal capacity at 0°C. The charging trigger decision refers to the decision of the robot to return to the base station for charging.

[0087] Specifically, the energy management module calculates the safe return power threshold and controls the margin coefficient to optimize charging timing while ensuring mission completion and safe return, thereby balancing operational continuity and battery health.

[0088] Specifically, the pesticide management module obtains the entropy of the remaining pesticide solution based on the predicted distribution of pest density. Specifically, it calculates the entropy of the remaining pesticide solution in the region. percentage To obtain, among which, Indicates the order of regions and predicts distribution based on pest density. and the proportion of the remaining drug solution in region i The entropy Y of the remaining liquid medicine is calculated, and Y is set as follows: ;

[0089] The remaining entropy of the medicine solution Y is compared with the preset remaining entropy of the medicine solution Y0. Based on the comparison result, the state of the remaining entropy of the medicine solution is judged, and the replenishment trigger decision is output based on the judgment result. Among them:

[0090] When Y≤Y0, the liquid management module determines that the state of the liquid remaining entropy is acceptable and does not output the liquid replenishment trigger decision;

[0091] When Y > Y0, the liquid management module determines that the state of the liquid remaining entropy is unacceptable, outputs the liquid replenishment trigger decision to the driver, and pushes the liquid replenishment trigger decision to the display screen.

[0092] Specifically, the remaining amount of the medicine solution is in the region The percentage refers to the percentage of hotel areas expected to be allocated at the current moment. The proportion of pesticide solution volume in each grid to the total remaining pesticide solution volume of the device is used in this embodiment. The proportion of spray volume in each area over the past 24 hours to the total spray volume is taken as the remaining pesticide solution volume in that area. The percentage of the remaining drug liquid entropy refers to the index that quantifies the uncertainty and urgency of the remaining drug liquid based on the information entropy theory. In this embodiment, the Shannon entropy formula is used to calculate the remaining drug liquid entropy. The preset remaining drug liquid entropy refers to the threshold for determining whether to trigger replenishment. In this embodiment, the function is solved according to the Shannon entropy formula and obtains a maximum value of approximately 0.693 when p=0.5. Therefore, Y0=0.693 is set. The state of the remaining drug liquid entropy refers to the acceptable degree of the remaining drug liquid entropy, including acceptable and unacceptable. The replenishment trigger decision refers to the decision of the robot to return to the base station for replenishment. This embodiment does not limit the specific method of pushing the charging trigger decision and the replenishment trigger decision to the display screen. Those skilled in the art can freely choose according to actual needs, such as pushing the charging trigger decision and the replenishment trigger decision to the display screen via Wi-Fi.

[0093] Specifically, the liquid medicine management module uses quantitative assessment of the liquid medicine balance entropy to trigger timely replenishment when the liquid medicine is insufficient, while avoiding invalid replenishment requests when the liquid medicine is close to full capacity, thereby optimizing the efficiency of liquid medicine management and storage utilization.

[0094] Specifically, the load management module obtains the charging urgency based on the safe return-to-home power threshold, specifically by calculating the charging urgency Uc based on the remaining power F2, the safe return-to-home power threshold F3, and the charging urgency weighting coefficient wc, and setting Uc=wc×(F3 / F2).

[0095] The load management module obtains the urgency of replenishment based on the drug liquid remaining entropy, and calculates the urgency of replenishment Uf based on the drug liquid remaining entropy Y, the preset drug liquid remaining entropy Y0 and the replenishment urgency weighting coefficient wf, and sets Uf=wf×(Y / Y0).

[0096] The load management module outputs a drug solution-charging urgency score based on the charging urgency and the replenishment urgency. Specifically, it calculates the drug solution-charging urgency score S based on the charging urgency Uc, the replenishment urgency Uf, the charging demand identifier Dc, and the replenishment demand identifier Df, and sets S = Dc × Uc + Df × Uf.

[0097] Specifically, the charging urgency refers to a quantitative indicator measuring the urgency of charging needs; a higher value indicates a more urgent charging need. Similarly, the replenishment urgency refers to a quantitative indicator measuring the urgency of replenishment needs; a higher value indicates a more urgent replenishment need. In this embodiment, a weighted simulation test is used to set the charging urgency weight coefficient wc=5 and the replenishment urgency weight coefficient wf=3. The weighted simulation test involves mapping the charging urgency weight coefficient and the replenishment urgency weight coefficient to the [0,10] interval, adjusting the weights, observing the system's decision-making behavior when both battery power and liquid solution are insufficient, and recording the replenishment urgency weight coefficient and charging urgency weight coefficient that result in the most stable system performance as the final values. The charging demand identifier refers to the numerical representation of the output of the charging trigger decision. The process of displaying the fluid replenishment requirement identifier refers to the process of numerically representing the output of the fluid replenishment trigger decision. This embodiment does not limit the specific method of numerical representation. Those skilled in the art can freely choose according to actual needs. For example, Dc=1 represents outputting a charging trigger decision, Dc=0 represents not outputting a charging trigger decision, Df=1 represents outputting a fluid replenishment trigger decision, and Df=0 represents not outputting a fluid replenishment trigger decision. The liquid-charging urgency score is an indicator that comprehensively evaluates the urgency of charging and fluid replenishment needs. The larger the value, the more urgent the need. The liquid-charging urgency score integrates a single need through a weighted summation method. When only charging is needed, S=Uc; when only fluid replenishment is needed, S=Uf; when both charging and fluid replenishment are needed, S=Uc+Uf.

[0098] Specifically, the load management module uses a comprehensive calculation of the urgency score between liquid and charging to uniformly quantify the urgency level when multiple service demands exist simultaneously, thereby supporting the comparability ranking of multi-dimensional demands.

[0099] Specifically, the load management module obtains the comprehensive load index of the port by: obtaining the candidate base station set Zi={Z1,Z2,...,Zn}, where n represents the order, n=1,2,...,n; and obtaining the charging port occupancy rate Ri, liquid replenishment port occupancy rate Hi, expected waiting time Ti, and distance Di of the candidate base station set Zi.

[0100] The port comprehensive load index Li is calculated based on the charging urgency Uc, the liquid replenishment urgency Uf, the charging port occupancy rate Ri, the liquid replenishment port occupancy rate Hi, the expected waiting time Ti, the distance Di, and the load weight set Wl={wl1,wl2,wl3,wl4}. The formula is: Li=wl1×(Ri×Uc+Li×Uf)+wl2×Ti+wl3×Di+wl4×(Uc+Uf);

[0101] The load management module optimizes the process of obtaining the port comprehensive load index based on the charging trigger decision and the liquid replenishment trigger decision, wherein:

[0102] When the output charging trigger decision is not outputting the liquid replenishment trigger decision, the process of obtaining the port comprehensive load index is optimized by weighting: the liquid replenishment urgency weight is reset to 0, Uf'=0 is set, and the port comprehensive load index Li is recalculated.

[0103] When no charging trigger decision is output but a liquid replenishment trigger decision is output, the process of obtaining the port comprehensive load index is optimized by weighting: the charging urgency weight is reset to 0, UC'=0 is set, and the port comprehensive load index Li is recalculated.

[0104] When the output charging trigger decision and the output liquid replenishment trigger decision are both made, the process of obtaining the port comprehensive load index is not optimized by weight.

[0105] Specifically, the candidate base station set refers to the set of stations within the device's communication range that can provide charging and liquid replenishment services, where Z1 represents the first base station, Z2 the second base station, and Zn the nth base station. The charging port occupancy rate refers to the proportion of base station charging ports that are occupied; in this embodiment, the charging port occupancy rate is obtained through base station status broadcast messages. The liquid replenishment port occupancy rate refers to the proportion of base station liquid replenishment ports that are occupied; in this embodiment, the liquid replenishment port occupancy rate Hi is obtained through base station status broadcast messages. The estimated waiting time refers to the estimated service time, in minutes, that the device needs to wait after arriving at the charging base station; in this embodiment, the estimated waiting time Ti is estimated using a queuing theory model. The distance refers to the distance between the device and the base station. The straight-line distance from the current location to the base station is in meters. In this embodiment, the distance is obtained using Euclidean distance. The load weight set Wl={wl1,wl2,wl3,wl4} includes port occupancy weight wl1, waiting time weight wl2, distance weight wl3, and urgency weight wl4. This embodiment does not limit the specific values ​​of wl1, wl2, wl3, and wl4. Those skilled in the art can freely choose according to actual needs. For example, by solving the Pareto optimal weight combination through a multi-objective optimization algorithm, wl1=0.3, wl2=0.2, wl3=0.3, and wl4=0.2 can be set. The port comprehensive load index refers to the comprehensive load index of a single base station.

[0106] Specifically, the load management module uses multi-factor modeling of the port comprehensive load index to comprehensively evaluate the service capabilities and waiting costs of multiple candidate base stations, thereby providing a quantitative basis for the selection of the optimal site.

[0107] Specifically, the load management module outputs the target base station based on the liquid-charging urgency score and the port comprehensive load index according to the principle of minimizing the maximum load. It selects the base station that minimizes max(Li) as the target base station based on the port comprehensive load index Li. When there are multiple base stations with the same max(Li), the base station with the shortest distance Di is selected as the target base station.

[0108] Specifically, the principle of minimizing maximum load refers to minimizing the load of the busiest node in the system by balancing the allocation of tasks in multi-robot scheduling, thereby avoiding local overload and global efficiency degradation. The max(Li) refers to the base station with the largest port comprehensive load index. In this embodiment, this principle is used to select the target base station. The target base station refers to the optimal service site selected for the current device according to the load balancing principle. In this embodiment, after outputting the target base station, the device autonomously navigates to the base station for charging and electrolyte replenishment.

[0109] Specifically, the load management module optimizes the scheduling based on the principle of minimizing the maximum load to achieve balanced charging load in multi-robot scenarios. This helps avoid port congestion when multiple robots share charging infrastructure, thereby achieving optimal global service efficiency and maximizing overall system throughput.

[0110] Please see Figure 3 As shown, this is a flowchart illustrating the task energy consumption prediction method of this embodiment. The task energy consumption prediction method includes:

[0111] Step S1: Map the task status data to the [0,1] interval using the maximum-minimum normalization method to obtain the standard value Sq of the area to be eliminated, the standard value Davg of the predicted pest density, the standard value Ctm of the terrain complexity, and the standard value Uem of the task urgency.

[0112] Step S2: Set the energy consumption weight set We={we1,we2,we3,we4};

[0113] Step S3: Calculate the expected energy consumption F1 of the task based on the standard value of the area to be exterminated Sq, the standard value of the predicted pest density Davg, the standard value of the terrain complexity Ctm, the standard value of the task urgency Uem, and the energy consumption weight set We={we1,we2,we3,we4}. Set F1=(Sq×Davg×we1+Ct×we2+Ue×we3)×we4, and we1+we2+we3=1.

[0114] Specifically, the energy consumption weight set We={we1,we2,we3,we4} includes pest density weight we1, terrain complexity weight we2, urgency weight we3, and energy consumption benchmark coefficient we4. The pest density weight is a coefficient that measures the importance of the standard value of the area to be eradicated and the standard value of the predicted pest density in the expected energy consumption of the task. The terrain complexity weight is a coefficient that measures the importance of the standard value of terrain complexity in the expected energy consumption of the task. The urgency weight is a coefficient that measures the importance of the standard value of the urgency of the task in the expected energy consumption of the task. The energy consumption benchmark coefficient is the coefficient used to convert the weighted comprehensive index into actual... The scaling factors for the expected energy consumption of the task are not limited in this embodiment. The specific values ​​of we1, we2, we3, and we4 are not limited. Those skilled in the art can freely choose them according to actual needs. For example, we1=0.5, we2=0.2, we3=0.3, and we4=0.8kWh / (m²·unit / m²) can be set by historical energy consumption data analysis. The historical energy consumption data analysis method refers to the process of taking the weight value that minimizes the overall error between the expected energy consumption of the task and the actual energy consumption in the past as the specific value of we1, we2, we3, and we4. The expected energy consumption of the task refers to the total amount of electricity expected to be consumed to complete the current task.

[0115] Specifically, the energy management module estimates task energy consumption through multi-factor comprehensive modeling, so as to provide a quantitative basis for charging decisions and avoid work interruption caused by insufficient power during task execution.

[0116] The technical solution of the present invention has been described above with reference to the preferred embodiments shown in the accompanying drawings. However, it will be readily understood by those skilled in the art that the scope of protection of the present invention is obviously not limited to these specific embodiments. Without departing from the principles of the present invention, those skilled in the art can make equivalent changes or substitutions to the relevant technical features, and the technical solutions after these changes or substitutions will all fall within the scope of protection of the present invention.

Claims

1. A hotel pest control robot capable of autonomous charging and pesticide management, characterized in that, include: The autonomous charging and pesticide management system, connected to the mobile chassis, is used to acquire pest environmental monitoring data and task status data, acquire pest density prediction distribution, acquire expected task energy consumption and pesticide residue entropy, and output charging trigger decisions and replenishment trigger decisions based on expected task energy consumption and pesticide residue entropy, acquire charging urgency and replenishment urgency, acquire pesticide-charging urgency score and port comprehensive load index, and output to the target base station according to the principle of minimizing maximum load; A sensor cluster, connected to a mobile chassis and an autonomous charging and pesticide management system, is used to collect pest environmental monitoring data. Atomizing nozzle, which is connected to the medicine tank, is used to spray the medicine. The charging port connects to the mobile chassis and is used to connect to the charging port for charging. The display screen, connected to the balancer, is used to display charging trigger decisions and replenishment trigger decisions; A drive unit, connected to the mobile chassis, is used to drive the mobile chassis to move based on charging trigger decisions and replenishment trigger decisions; The mobile chassis, which is connected to the autonomous charging and medicine management system, sensor cluster, charging port, driver, balancer and medicine tank, is used to carry the device for movement; A balancer, which connects to the mobile chassis and display screen, is used to maintain balance; The medicine tank, which is connected to the mobile chassis and atomizing nozzle, is used to store the medicine.

2. The hotel pest control robot capable of autonomous charging and pesticide management according to claim 1, characterized in that, The autonomous charging and medication management system includes: The environmental sensing module is used to acquire pest environmental monitoring data and task status data; The pest prediction module is used to extract features from pest environmental monitoring data to obtain pest environmental feature vectors. It is also used to perform model recognition based on pest environmental feature vectors to obtain the predicted distribution of pest density. The energy management module is used to acquire task status data, estimate task energy consumption based on the task status data, obtain the expected energy consumption of the task, acquire the remaining power and the safe return power threshold, and output the charging trigger decision based on the expected energy consumption, remaining power and safe return power threshold. The pesticide management module is used to obtain the pesticide residue entropy based on the predicted distribution of pest density, and output the replenishment trigger decision based on the pesticide residue entropy. The load management module is used to obtain the charging urgency based on the safe return-to-home power threshold, the replenishment urgency based on the liquid level entropy, and to obtain a liquid-charging urgency score based on both the charging urgency and replenishment urgency. It is also used to obtain the port comprehensive load index and to optimize the acquisition process of the port comprehensive load index based on the charging trigger decision and the replenishment trigger decision. Finally, it outputs the target base station based on the liquid-charging urgency score and the port comprehensive load index by minimizing the maximum load principle.

3. The hotel pest control robot capable of autonomous charging and pesticide management according to claim 2, characterized in that, The pest prediction module extracts features from pest environmental monitoring data, specifically by: normalizing temperature and humidity data to obtain normalized ambient temperature T' and normalized relative humidity H'; performing a logarithmic transformation on illuminance data L to obtain transformed illuminance data L', setting L'=ln(L+1); standardizing carbon dioxide concentration data C to obtain standardized carbon dioxide concentration data C'; encoding and converting human presence data P to obtain encoded human presence data P'; and concatenating the normalized ambient temperature T', normalized relative humidity H', transformed illuminance data L', standardized carbon dioxide concentration data C', and encoded human presence data P' to obtain the pest environmental feature vector E={T',H',L',C',P'}.

4. The hotel pest control robot capable of autonomous charging and pesticide management according to claim 3, characterized in that, The pest prediction module performs model recognition based on pest environmental feature vectors. Specifically, it constructs a pest density prediction model by inputting the pest environmental feature vector E into the model to obtain the predicted pest density distribution output by the model. ={ , ,..., },in, Indicates the order of regions. =1,2,……, , The total number of grid units for the hotel area.

5. The hotel pest control robot capable of autonomous charging and pesticide management according to claim 4, characterized in that, The energy management module acquires task status data and estimates task energy consumption based on the task status data. Specifically, the task status data is mapped to the [0,1] interval by the maximum-minimum normalization method to obtain the standard value Sq of the area to be eliminated, the standard value Davg of the predicted pest density, the standard value Ctm of the terrain complexity, and the standard value Uem of the task urgency. Based on the standard value Sq of the area to be exterminated, the standard value Davg of the predicted pest density, the standard value Ctm of the terrain complexity, the standard value Uem of the task urgency, and the energy consumption weight set We={we1,we2,we3,we4}, the expected energy consumption F1 of the task is calculated, and F1 is set as (Sq×Davg×we1+Ct×we2+Ue×we3)×we4, and we1+we2+we3=1.

6. The hotel pest control robot capable of autonomous charging and pesticide management according to claim 5, characterized in that, The energy management module acquires the remaining battery power and the safe return battery power threshold, specifically: it acquires the remaining battery power F2 in real time through the battery management system, and determines the threshold based on the device's current location coordinates ( , ), the coordinates of the nearest base station ( , Energy consumption per unit distance for return and return safety factor The safe return-to-home battery threshold F3 is calculated and set as F3= , and 1 < <2; The energy management module outputs a charging trigger decision based on the expected energy consumption F1, remaining power F2, and safe return power threshold F3, wherein: When F1+F3<F2×η1, no output is made for the charging trigger decision; When F1+F3≥F2×η1, the charging trigger decision is output to the driver, where η1 is the power safety margin coefficient and 0<η1<1, and the charging trigger decision is pushed to the display screen.

7. The hotel pest control robot capable of autonomous charging and pesticide management according to claim 6, characterized in that, The pesticide management module obtains the entropy of the remaining pesticide solution based on the predicted distribution of pest density. Specifically, it calculates the entropy of the remaining pesticide solution in the region. percentage To obtain, among which, Indicates the order of regions and predicts distribution based on pest density. and the proportion of the remaining drug solution in region i The entropy Y of the remaining liquid medicine is calculated, and Y is set as follows: ; The remaining entropy of the medicine solution Y is compared with the preset remaining entropy of the medicine solution Y0. Based on the comparison result, the state of the remaining entropy of the medicine solution is judged, and the replenishment trigger decision is output based on the judgment result. Among them: When Y≤Y0, the liquid management module determines that the state of the liquid remaining entropy is acceptable and does not output the liquid replenishment trigger decision; When Y > Y0, the liquid management module determines that the state of the liquid remaining entropy is unacceptable, outputs the liquid replenishment trigger decision to the driver, and pushes the liquid replenishment trigger decision to the display screen.

8. The hotel pest control robot capable of autonomous charging and pesticide management according to claim 7, characterized in that, The load management module obtains the charging urgency based on the safe return power threshold. Specifically, it calculates the charging urgency Uc based on the remaining power F2, the safe return power threshold F3, and the charging urgency weighting coefficient wc, and sets Uc=wc×(F3 / F2). The load management module obtains the urgency of replenishment based on the drug liquid remaining entropy, and calculates the urgency of replenishment Uf based on the drug liquid remaining entropy Y, the preset drug liquid remaining entropy Y0 and the replenishment urgency weighting coefficient wf, and sets Uf=wf×(Y / Y0). The load management module outputs a drug solution-charging urgency score based on the charging urgency and the replenishment urgency. Specifically, it calculates the drug solution-charging urgency score S based on the charging urgency Uc, the replenishment urgency Uf, the charging demand identifier Dc, and the replenishment demand identifier Df, and sets S = Dc × Uc + Df × Uf.

9. The hotel pest control robot capable of autonomous charging and pesticide management according to claim 8, characterized in that, The load management module obtains the comprehensive load index of the ports, specifically by: obtaining the candidate base station set Zi={Z1,Z2,...,Zn}, where n represents the order, n=1,2,...,n; and obtaining the charging port occupancy rate Ri, liquid replenishment port occupancy rate Hi, estimated waiting time Ti, and distance Di of the candidate base station set Zi. The port comprehensive load index Li is calculated based on the charging urgency Uc, the liquid replenishment urgency Uf, the charging port occupancy rate Ri, the liquid replenishment port occupancy rate Hi, the expected waiting time Ti, the distance Di, and the load weight set Wl={wl1,wl2,wl3,wl4}. The formula is: Li=wl1×(Ri×Uc+Li×Uf)+wl2×Ti+wl3×Di+wl4×(Uc+Uf); The load management module optimizes the process of obtaining the port comprehensive load index based on the charging trigger decision and the liquid replenishment trigger decision, wherein: When the output charging trigger decision is not outputting the liquid replenishment trigger decision, the process of obtaining the port comprehensive load index is optimized by weighting: the liquid replenishment urgency weight is reset to 0, Uf'=0 is set, and the port comprehensive load index Li is recalculated. When no charging trigger decision is output but a liquid replenishment trigger decision is output, the process of obtaining the port comprehensive load index is optimized by weighting: the charging urgency weight is reset to 0, UC'=0 is set, and the port comprehensive load index Li is recalculated. When the output charging trigger decision and the output liquid replenishment trigger decision are both made, the process of obtaining the port comprehensive load index is not optimized by weight.

10. The hotel pest control robot capable of autonomous charging and pesticide management according to claim 8, characterized in that, The load management module outputs the target base station based on the liquid-charging urgency score and the port comprehensive load index according to the principle of minimizing the maximum load. It selects the base station that minimizes max(Li) as the target base station based on the port comprehensive load index Li. When there are multiple base stations with the same max(Li), the base station with the shortest distance Di is selected as the target base station.

Images