Central air conditioning pipeline intelligent cleaning robot based on whale optimization algorithm

Abstract

This invention relates to the field of intelligent cleaning equipment technology, specifically to an intelligent cleaning robot for central air conditioning ducts based on the Whale Optimization Algorithm. The cleaning process is synchronized with the moving mechanism: a rotating brush sweeps the top and sides of the duct at high speed, a strip brush scrapes away residual dust on the duct walls as the vehicle moves, a rotating scraper at the bottom concentrates dust to the center, and a negative pressure suction device at the rear adsorbs dust in real time, forming a closed-loop cleaning process of "sweeping-concentration-adsorption" to prevent secondary dust diffusion. The beneficial effects of this invention are: 1. Safety and reliability: It replaces manual high-altitude work, reducing the accident rate from 32% to below 5%, preventing workers from inhaling pollutants, and ensuring operational safety and health; 2. High cleaning efficiency: The WOA algorithm achieves a cleaning coverage rate of over 98.5%, increasing the cleanliness of curved pipe areas from 65% to 95%, and shortening the single operation time to less than 4 hours, which is 4 times more efficient than manual cleaning.

Description

Technical Field

[0001] This invention relates to the field of intelligent cleaning equipment technology, specifically to an intelligent cleaning robot for central air conditioning ducts based on the whale optimization algorithm. Background Technology

[0002] With the acceleration of urbanization, central air conditioning systems have become standard equipment in modern buildings. However, the long-term accumulation of pollutants such as PM2.5 particles, Legionella, and Aspergillus inside the ducts not only threatens public health, leading to 37.2% of sick building syndrome cases, but also increases the drag coefficient by 18-25%, causing a 30% increase in fan energy consumption. Traditional manual cleaning has many drawbacks: the accident rate for high-altitude operations reaches 32%, posing a high health risk; single downtime is ≥8 hours, resulting in low efficiency; the cleanliness of bends is only 65%, leaving many blind spots; and it requires supporting protective equipment, leading to high maintenance costs.

[0003] Existing pipeline cleaning robots mostly use static path presets or remote manual control, rely on prior maps, and have a lagging dynamic response. When faced with complex pipeline network topologies, traditional algorithms (such as ant colony algorithms and particle swarm algorithms) have high parameter sensitivity and strong local convergence, making it difficult to achieve fully autonomous coverage navigation in a three-dimensional narrow space. The coverage rate is less than 85%, and frequent manual intervention is still required.

[0004] Therefore, developing a central air conditioning duct cleaning robot with intelligent path planning, strong environmental adaptability, and high cleaning efficiency has become a key requirement for solving the bottleneck of building air safety. Summary of the Invention

[0005] This invention aims to provide an intelligent cleaning robot for central air conditioning ducts based on the whale optimization algorithm, which solves the technical problems of traditional manual cleaning, such as high safety hazards, low efficiency, and many blind spots, as well as the rigid path planning and poor adaptability of existing robots, so as to achieve safe, efficient and full-coverage cleaning of central air conditioning ducts.

[0006] The core technical solution of this invention is as follows:

[0007] Mechanical structure module

[0008] Wheeled mobile chassis: Adopting a three-wheel, two-drive structure, the front wheels are flat-diameter rubber wheels driven by two TT DC micro-gear motors, and the rear wheels are solid nylon universal wheels. It supports forward, backward, turning, and stationary rotation, with a minimum turning radius ≤0.3m. The chassis is made of acrylic sheet material, with overall dimensions of 350mm × (290mm~390mm) × (290mm~310mm) and a total weight ≤8kg. It is compatible with rectangular pipes of different sizes.

[0009] The adaptive telescopic cleaning mechanism consists of three parts: ① Two nylon rotating brushes at the front of the vehicle are positioned at a 45° angle to the horizontal plane and are driven by a 130 DC motor (speed ≥2000rpm) to perform initial cleaning of the top and sides of the duct; ② Nylon strip brushes are installed on the sides and top of the vehicle. The side strip brushes are connected by a slider telescopic structure simulated by a syringe, and their width can be adjusted within the range of 290mm~390mm. The height of the top strip brush can be adjusted by changing its size, achieving multi-faceted cleaning of the duct without dead angles; ③ A servo-driven rotating scraper is installed at the bottom. When the machine is stopped, it is in a figure-eight shape. After starting, it rotates from the outside to the inside to a parallel state, circulating and concentrating the dust to the center of the vehicle.

[0010] Negative pressure dust collection device: Installed at the rear of the vehicle, it consists of a geared motor, a small fan blade, a filter screen and a circular dust collection nozzle. The motor drives the fan blade to rotate at high speed to generate negative pressure, which adsorbs the concentrated dust. The filter screen prevents dust from flowing back, and the nozzle is designed to adapt to the bottom cleaning trajectory.

[0011] Hardware control module

[0012] Core control unit: Based on the STM32F469 microcontroller (ARM Cortex-M4 core, 168MHz main frequency), it is responsible for receiving sensor data, generating PWM drive signals, and controlling the actuators; the Raspberry Pi 4B coprocessor (equipped with R7FA4M1AB3CFM main controller) runs the WOA path planning algorithm, receives camera images through the CSI interface, constructs a pipeline environment point cloud map in combination with SLAM, and sends path commands to the STM32F469 through the CAN bus (communication cycle ≤10ms).

[0013] Sensor Units: Ultrasonic ranging module (HC-SR04) is used to detect the distance to obstacles with an accuracy of ±1mm; Infrared obstacle avoidance module (LM393) has a response time of <10ms and outputs high and low level signals to identify obstacles; Environmental sensors collect PM2.5, temperature, humidity and VOC data to provide a basis for cleaning efficiency assessment.

[0014] Drive unit: The L298N drive module is used, with a maximum operating voltage of 46V and a maximum output current of 4A. It supports forward and reverse rotation of the motor and stepless speed regulation, and has a built-in automatic current limiting protection function to ensure stable system operation.

[0015] Path planning module

[0016] Path planning is implemented based on the Whale Optimization Algorithm (WOA), with a deep mapping between the algorithm and hardware functionality:

[0017] Hunting Prey (Global Search): Initialize the whale pod with Raspberry Pi 4B, starting from the pipe inlet and taking the uncleaned area as the prey, and generate a full-coverage path point sequence (X,Y,Z coordinates).

[0018] Bubble web predation (partial development): STM32F469 uses a spiral equation to calculate the relative angle between the real-time position and the target point. Calculate the motor differential ratio and approach the target cleaning point along the spiral path.

[0019] Searching for prey (random exploration): When the infrared sensor detects an unknown obstacle, a random search mode is triggered. A path around the obstacle is generated through WOA, and combined with the obstacle penalty function (the fitness value increases sharply when a candidate path enters the obstacle grid), the whale pod is forced to move away from the obstacle area.

[0020] Meanwhile, a two-dimensional gridded model of the pipeline is constructed, and a grid continuity penalty term is added to the WOA fitness function. When the adjacent path points are not in a 4-neighborhood, λ is set to 1000 to force path continuity. The directional weights are dynamically adjusted, with the weight of "downward" movement increased in the upper-level region and the weight of "rightward" movement increased in the lower-level region to avoid local optima.

[0021] Cleaning Execution Module

[0022] The cleaning process starts simultaneously with the moving mechanism: the rotating brush sweeps the top and sides of the air duct at high speed, the strip brush scrapes the dust remaining on the duct wall as the vehicle moves, the bottom rotating scraper concentrates the dust to the center, and the negative pressure dust suction device at the rear adsorbs the dust in real time, forming a closed-loop cleaning process of "sweeping-concentration-adsorption" to avoid secondary diffusion of dust.

[0023] The beneficial effects of this invention are:

[0024] Safe and reliable: It replaces manual high-altitude work, reducing the accident rate from 32% to below 5%, avoiding workers' inhalation of pollutants, and ensuring work safety and health.

[0025] Clean and efficient: The WOA algorithm achieves a cleaning coverage rate of over 98.5%, increases the cleanliness of the bend area from 65% to 95%, and shortens the single operation time to less than 4 hours, which is 4 times more efficient than manual cleaning.

[0026] Highly adaptable: The telescopic cleaning mechanism can adapt to rectangular pipes ranging from 200×200mm to 600×400mm, meeting the cleaning needs of ducts of different sizes without changing equipment, thus solving the size adaptation limitations of traditional robots.

[0027] Energy-saving and environmentally friendly: It is powered by electricity, has no exhaust emissions, reduces the use of chemical cleaning agents by 60%, and reduces building energy consumption by 8%-12%, which meets the requirements of the "dual carbon" strategy.

[0028] Intelligent and controllable: Equipped with environmental sensors and cloud communication capabilities, it generates real-time cleaning performance reports, supports remote monitoring and parameter optimization, and enables intelligent operation and maintenance.

[0029] Obviously, based on the above description of the present invention, and according to common technical knowledge and conventional methods in the field, various other modifications, substitutions or alterations can be made without departing from the basic technical concept of the present invention.

[0030] The following detailed embodiments further illustrate the above-described content of the present invention. However, this should not be construed as limiting the scope of the present invention to the following examples. All technologies implemented based on the above-described content of the present invention fall within the scope of the present invention. Attached Figure Description

[0031] Figure 1 This is a schematic diagram of the main structure of the mechanical structure module of the present invention;

[0032] Figure 2 This is a right-side structural schematic diagram of the mechanical structure module of the present invention;

[0033] Figure 3 This is a bottom view of the mechanical structure module of the present invention;

[0034] Figure 4 This is a slanted view of the mechanical structure module of the present invention;

[0035] Figure 5 This is a schematic diagram of the wiring structure of the L298N driver module of the present invention;

[0036] Figure 6 This is a schematic diagram of the system control flow structure of the present invention;

[0037] Figure 7 This is a schematic diagram of the optimal Whale Path of Exploration (WOA) structure of the present invention.

[0038] Reference table for attached figures:

[0039] 1. Vehicle body; 2. Rubber wheels; 3. Casters; 4. Negative pressure dust collection device; 5. Rotating brush; 6. Ultrasonic rangefinder; 7. Rotating scraper; 8. Camera; 9. Telescopic strip brush. Detailed Implementation

[0040] The present invention is illustrated below with specific embodiments, but these are not intended to limit the scope of the invention. Figure 1-7 As shown, a smart cleaning robot for central air conditioning ducts based on the whale optimization algorithm is presented.

[0041] Example 1

[0042] Mechanical structure assembly

[0043] The mechanical structure module includes a vehicle body 1. Rubber wheels 2 are provided on both sides of the downward end of the vehicle body 1 for movement. Universal wheels 3 are provided on both sides of the downward end of the vehicle body 1 away from the rubber wheels 2 for movement. An ultrasonic rangefinder 6 is provided at one end of the vehicle body 1. A camera 8 is provided at the top of the vehicle body 1. A rotating brush 5 is provided at the end of the vehicle body 1 near the ultrasonic rangefinder 6. A rotating scraper 7 is provided downward on the vehicle body 1. Telescopic strip brushes 9 are provided on both sides of the vehicle body 1. A negative pressure dust collection device 4 is provided at the end of the vehicle body 1 away from the ultrasonic rangefinder 6.

[0044] The negative pressure dust collection device 4 is used to collect the dust and dirt swept down by the rotating brush, preventing it from being scattered back into the pipe.

[0045] The chassis is made of 5mm thick acrylic sheet, with dimensions of 350mm×390mm×310mm. The front wheel is equipped with a TT DC micro geared motor (dual-shaft type, rated power 0.625W, torque 1.75N・m), and the rear wheel is fixed with nylon universal wheels. The wheel track is set to 290mm to ensure turning flexibility.

[0046] The rotating brush at the front of the vehicle has a diameter of 70mm and an angle of 45° with the horizontal plane. It is connected to a 130 DC motor (78.5W power, 1500rpm speed) via a small iron strip. The motor is fixed to the brackets on both sides of the front of the vehicle. The side strip brushes are 310mm long and are connected via a telescopic slider with a telescopic stroke of 100mm. The top strip brush is 350mm long and is fixed to the top of the vehicle body with bolts. Different height strip brushes (290mm~310mm) can be replaced. The bottom servo motor (model SG90) is fixed to the center of the chassis and connected to the figure-eight scraper. The servo motor has a rotation range of 0°~90°.

[0047] The negative pressure dust collection device uses a cylindrical plastic bottle (50mm in diameter and 150mm in length), with a built-in 12V geared motor and fan blades. A circular dust collection nozzle (30mm in diameter) is installed at the bottle opening, and a filter screen is installed inside the bottle. It is connected to the pins of the secondary control board via DuPont wires.

[0048] Hardware connection

[0049] The PA0-PA7 pins of the STM32F469 microcontroller are connected to the OUT pins of the infrared obstacle avoidance module, and the PB6-PB7 pins are connected to the ultrasonic ranging module via the I2C bus. The TIM1_CH1-4 pins output PWM signals to the L298N driver module to control the front wheel motor and the cleaning motor. The CSI interface of the Raspberry Pi 4B is connected to the OV5647 camera, the CAN_H / CAN_L pins communicate with the STM32F469, and the WiFi module enables data transmission with the cloud platform.

[0050] Infrared obstacle avoidance modules are common sensor modules, primarily based on infrared ranging sensors, especially utilizing the principle of triangulation, to detect the presence and distance of objects. They operate by emitting infrared light and receiving the reflected light, thus achieving obstacle avoidance. An infrared obstacle avoidance module typically includes an infrared transmitter and an infrared receiver. The transmitter emits infrared light of a specific wavelength; when the light encounters an obstacle, it is reflected back and captured by the receiver. By measuring the intensity and return time of the light, the distance and position of the obstacle can be determined. The working principle of infrared obstacle avoidance sensors is mainly based on the reflection characteristics of infrared light (light reflection). These sensors emit infrared signals into the surrounding environment; when they encounter an obstacle, some of the infrared light is reflected back and received by the sensor's receiver. The received reflected signal generates a corresponding voltage signal; the intensity and angle of this signal are used by the sensor to determine the presence of an obstacle and its approximate position and distance. Such modules typically include an infrared transmitter and an infrared receiver. When the module is powered on, the infrared transmitter continuously emits infrared light of a specific frequency; this infrared light is reflected back when it encounters an obstacle and received by the receiver. The output signal (level change) of the receiver tube is used to determine whether there is an obstacle. If there is an obstacle in front, the reflected infrared light is received by the receiver tube, and a low-level signal is output at this time; if there is no obstacle in front, the infrared light is not reflected, and a high-level signal is output.

[0051] The L298N driver module operates primarily based on its dual H-bridge circuit structure, used to control the rotation direction and speed of DC motors or stepper motors. The L298N motor driver module is based on a full-bridge drive circuit, capable of controlling the forward and reverse rotation of DC motors and the rotation of stepper motors. It can be understood as the motor driver making the entire circuit more powerful. The motor driver module is a crucial component in controlling motor rotation, and the L298N is one commonly used type. Its actual onboard voltage drop is two volts, so the resulting voltage will be two volts lower. For example, if the motor used is a 5V motor, a 7V power supply should be used.

[0052] The power module uses four 1.5V batteries connected in series (total voltage 6V) to power the microcontroller and sensors, and outputs 5V to power the geared motor and servo motor through a voltage converter.

[0053] Algorithm Implementation

[0054] The WOA algorithm was implemented in Python on a Raspberry Pi 4B. The initial parameters were: number of whales 30, maximum number of iterations G=100, encirclement contraction coefficient a decreasing linearly from 2 to 0, strategy selection probability p∈[0,1], and spiral shape parameter b=1.

[0055] Pipeline grid modeling uses 20mm×20mm grid units. Inputting the pipe dimensions (such as 350mm×310mm) automatically generates a grid map. Black grids mark fixed obstacles such as pipe bends and joints.

[0056] The fitness function is set to F = cleaning coverage - 0.3 × operation time + 0.2 × energy consumption, which combines obstacle penalty and grid continuity penalty to ensure optimal path and compliance.

[0057] Power supply module: First, it provides power support for the entire system, meeting the dual independent power supply requirements of the self-propelled vehicle's power system and computing unit, and ensuring the stability of the system under dynamic loads.

[0058] The STM32F469, based on an ARM Cortex-M4 core (168MHz clock speed, FPU acceleration), acquires motor speed / steering angle data in real time via PWM encoding and drives the motion system (motor, servo). It integrates WOA algorithm-based dynamic path instructions to generate PWM waveforms with adjustable duty cycles (1ns resolution), enabling precise motion control of the self-propelled vehicle within the pipeline.

[0059] Raspberry Pi 4B Coprocessor: Equipped with an R7FA4M1AB3CFM main controller (48MHz Cortex-M4) and an ESP32-S3 dual-mode communication module, it runs a WOA-based 3D path planning algorithm, receives camera images (30fps@720p) through the CSI interface, constructs a pipeline environment point cloud map using SLAM, and sends real-time path commands to the STM32 via the CAN bus (communication cycle ≤10ms).

[0060] Camera: The camera is responsible for capturing image data and transmitting it to the Raspberry Pi 4B via the CSI interface.

[0061] QT host computer: Based on TCP / IP protocol, it realizes human-computer interaction, visualizes the heat map of pipeline cleaning coverage, WOA algorithm iteration curve and energy consumption data, and supports parameter optimization in manual intervention mode (such as dynamic correction of shrinkage coefficient a).

[0062] The Raspberry Pi 4 Model B uses the Sapphire RA4M1 (Arm® Cortex®-M4) main controller with a 48MHz clock speed. It features an onboard ESP32-S3 module, integrating 2.4GHz Wi-Fi and Bluetooth Low Energy (Bluetooth LE) for dual-mode wireless communication. The R7FA4 has 256kB Flash and 32kB RAM, capable of handling complex projects. The R7FA4PLUSB adds a 12-bit DAC, CAN bus, and operational amplifier, providing expanded functionality and flexibility for the design. It has an onboard ICSP interface, which can be used as an SPI interface, and also features a Qwiic connector for easy connection of I2C devices within the Qwiic ecosystem. The onboard ESP32-S3FN8 module has 384kB ROM, 512kB RAM, and 8MB Flash, providing ample space. The development board has 6 PWM control interfaces, 1 I2C interface, 1 SPI interface, and 1 CAN bus interface.

[0063] Operation process

[0064] After the robot is started, the sensor module initializes and collects initial environmental data of the pipeline. The Raspberry Pi 4B uses the camera and SLAM to build a point cloud map of the pipeline and runs the WOA algorithm to generate a global cleaning path.

[0065] The STM32F469 receives path instructions, drives the front wheel motor to move the robot along the planned path, and simultaneously starts the cleaning mechanism: the rotating brush and strip brush work synchronously, the bottom servo motor controls the scraper to clean in a cycle, and the tail vacuum cleaner starts to generate negative pressure.

[0066] During operation, the ultrasonic ranging module detects the distance to obstacles in real time, the infrared obstacle avoidance module identifies sudden obstacles, triggers the WOA algorithm random search mode to generate an obstacle avoidance path, and ensures collision-free operation; environmental sensors collect data in real time and transmit it to the QT host computer via WiFi to display the cleaning progress and environmental parameters.

[0067] After cleaning is completed, the robot returns to the pipe inlet along the original path and generates a cleaning efficiency report, which includes data such as cleaning coverage, cleanliness, and energy consumption.

[0068] Example 2

[0069] Cleaning solution:

[0070] Option 1: Install rotating roller brushes on the left and right sides of the front of the cart to knock off the dust on the walls of the air duct. Install a small dustpan at the rear of the cart and then use a sponge to clean up the dustpan.

[0071] Option 2: Install horizontal brushes at the front of the cart. The brushes are driven by a motor to rotate and knock off the dust from the walls of the air duct. Install a dust sweeper with a linkage structure at the bottom of the cart to sweep the dust that has fallen to the ground to the middle of the cart as it travels. Add a vacuum device at the rear to collect the dust on the top of the cart. This way, the dust can be sucked up from the rear after the cart has passed.

[0072] Option 3: Based on Option 2, add a rotating guide rod mechanism (crank rod) at the rear. After the dust is brushed off, the dust is collected at the bottom of the trolley and sucked away by the vacuum device at the rear. At this point, the basic cleaning of the air duct is basically completed. Then, the sponge installed on the rotating guide rod mechanism at the rear performs further cleaning of the four walls of the air duct by rotating.

[0073] Option 4: The dust removal design is divided into four parts. A roller brush is still used at the front of the cart to remove dust. Scrapers are installed on the left, right, and top sides of the cart to scrape the dust directly from all four sides, making dust removal more complete. At the bottom, a linkage structure is used to scrape the dust, causing the dust to concentrate in the center of the cart's path. Finally, a vacuum cleaner at the rear of the cart sucks up the dust and collects it in the dust collector on the top of the cart. This completes the cleaning of the self-propelled cart inside the duct.

[0074] Option 5: The dust removal design is divided into four parts. The first part uses roller brushes at the front of the vehicle to remove dust, with the two roller brushes at a 45° angle to the horizontal plane of the front of the vehicle. This can clean the top and sides of the air duct, removing dust from the three sides inside the air duct. In addition, to ensure the cleaning effect, scraper brushes are installed on the left and right sides and the top of the vehicle to scrape the dust directly from the top three sides, making the cleaning of the air duct more thorough. At the bottom, the rotation angle of the servo motor is controlled by microcontroller code, and brushes are installed on the servo motors. When the system is stopped, the brushes are in a figure-eight shape. After the system starts, the brushes can sweep inward from the figure-eight shape until the two brushes are in a parallel state, which can sweep the pollutants at the bottom to the center of the vehicle's travel path. Then, the vacuum cleaner at the rear sucks in the pollutants and dust at the bottom, thereby achieving the effect of cleaning the entire interior of the ventilation duct.

[0075] It should be noted that, in this document, relational terms such as "first" and "second" are used only 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 process, method, article, or apparatus.

[0076] The above description is merely a preferred embodiment of the present invention and does not limit the patent scope of the present invention. Any equivalent structural transformations made using the contents of the present invention's specification and drawings under the inventive concept of the present invention, or direct / indirect applications in other related technical fields, are included within the patent protection scope of the present invention.

[0077] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A central air conditioning duct intelligent cleaning robot based on the whale optimization algorithm, characterized in that, Mechanical structure module, hardware control module, path planning module, and cleaning execution module; The mechanical structure module includes a vehicle body (1), with rubber wheels (2) on both sides of the lower end of the vehicle body (1) for movement, and omnidirectional wheels (3) on both sides of the lower end of the vehicle body (1) away from the rubber wheels (2). An ultrasonic rangefinder (6) is provided at one end of the vehicle body (1), and a camera (8) is provided at the top of the vehicle body (1). A rotating brush (5) is provided at the end of the vehicle body (1) near the ultrasonic rangefinder (6), and a rotating scraper (7) is provided at the lower end of the vehicle body (1). Telescopic strip brushes (9) are provided on both sides of the vehicle body (1), and a negative pressure dust collection device (4) is provided at the end of the vehicle body (1) away from the ultrasonic rangefinder (6). The hardware control module is located inside the vehicle body (1) and includes a core control unit, a sensor unit, and a drive unit. The core control unit consists of an STM32F469 microcontroller and a Raspberry Pi 4B coprocessor. The STM32F469 microcontroller is responsible for motor drive, sensor data acquisition, and actuator control. The Raspberry Pi 4B coprocessor runs the WOA algorithm and constructs a pipeline environment point cloud map. The sensor unit includes an ultrasonic ranging module (HC-SR04), an infrared obstacle avoidance module (LM393), and environmental sensors (PM2.5 / temperature and humidity / VOC). The drive unit uses an L298N drive module to realize motor forward and reverse rotation and speed adjustment. The path planning module is located inside the vehicle body (1). Based on the whale optimization algorithm, it generates the globally optimal cleaning path by mapping the three-stage behavior of "hunting prey - bubble net predation - searching for prey" and combining pipeline grid modeling and obstacle penalty function. The hunting prey stage initializes the whale pod and generates a full-coverage path point sequence starting from the pipeline entrance. The bubble net predation stage calculates the motor differential ratio through the spiral equation to approach the target cleaning point. The searching for prey stage triggers the generation of random obstacle avoidance paths through infrared sensor feedback. The cleaning execution module is located inside the vehicle body (1). The rotating brush rotates at a speed of ≥2000 rpm. The bottom scraper is controlled by a servo motor to achieve a figure-eight shape to a parallel state for cyclic cleaning. The negative pressure dust collection device generates negative pressure to adsorb and concentrate dust.

2. The intelligent cleaning robot for central air conditioning ducts according to claim 1, characterized in that, The path planning process of the WOA algorithm includes: Step 1: Construct a two-dimensional gridded environment model of the pipeline, discretizing the inner wall of the pipeline into grid cells, with black grid cells representing obstacles and white grid cells representing passable areas; Step 2: Initialize whale pod parameters, including the number of individual whales, maximum number of iterations G, encirclement contraction coefficient a, strategy selection probability p, and spiral shape parameter b; Step 3: Calculate the fitness value for each individual whale; Step 4: When p≤0.5, execute the spiral attack strategy and update the path through the equation; when p>0.5 and |A|<1, execute the prey encirclement strategy; when p>0.5 and |A|≥1, execute the random search strategy. Step 5: Iterate and update to the maximum number of iterations G, output the optimal cleaning path, and transmit it to the STM32F469 microcontroller via the CAN bus.

3. The intelligent cleaning robot for central air conditioning ducts according to claim 1, characterized in that, The hardware control module is connected as follows: the PA0-PA7 pins of the STM32F469 are connected to the infrared obstacle avoidance sensor, the PB6-PB7 pins are connected to the ultrasonic ranging module via the I2C bus, and the TIM1_CH1-4 pins output PWM signals to the L298N driver module; the Raspberry Pi 4B is connected to the camera via the CSI interface, communicates with the STM32F469 via the CAN bus, and interacts with the QT host computer via the TCP / IP protocol.

Images