A water self-adaptive adjusting control system and method of a scrubber

Abstract

The present application relates to the technical field of process control, in particular to a kind of self-adaptive adjustment control system and method of water quantity of floor cleaning machine, including dirt identification module, double-channel collaborative decision module and double-channel control execution module.The present application realizes once perception, simultaneously outputs the double-dimension information of dirt type and dirt range, and realizes the simultaneous approximation optimum of cleaning efficiency and water saving and energy saving, makes intensity and water quantity can be independently optimized to respective optimum value, avoids the dilemma of traditional coupling control, forced to accept waste for improving efficiency or forced to sacrifice effect for saving resources, simultaneously, through real-time feedback regulation module, realizes the dynamic termination of cleaning result orientation instead of time-oriented fixed length cleaning, avoids the cleaning bottleneck caused by dry friction and the water stain residue caused by excessive water supply, and under the premise of guaranteeing cleaning effect, maximum degree saves water and electricity resources and reduces mechanical wear.

Description

Technical Field

[0001] This invention relates to the field of process control technology, and more specifically, to a floor scrubber water volume adaptive adjustment control system and method. Background Technology

[0002] Floor scrubbers, as a type of floor cleaning equipment that integrates sweeping, mopping, and vacuuming functions, have been widely used in household and commercial cleaning scenarios. During operation, the control of water output and water pressure is a key factor in determining the cleaning effect, the amount of residual water on the floor, and the equipment's battery life.

[0003] Existing technology uses turbidity detection devices to detect the degree of dirtiness in the sewage in the sewage recovery pipe, thereby determining the amount of water needed to clean the dirt and controlling the water flow. However, it cannot distinguish the type and spatial distribution of dirt. Different types of dirt (such as stubborn adhesive stains and large-area splashes) have completely different requirements for cleaning power. Controlling the water flow based on a single parameter of dirt level makes it difficult to achieve precise cleaning. Furthermore, relying on user experience to judge the degree of dirt and manually switch water levels is not only cumbersome to operate, but also cannot be adjusted in real time according to the dynamic changes of dirt on the ground, which can easily lead to incomplete cleaning or waste of water resources. In view of this, we propose a floor scrubbing machine water volume adaptive adjustment control system and method that can achieve accurate dual-dimensional identification of dirt type and range, has self-verification and error correction capabilities for identification results, and can independently optimize water output intensity and volume based on dirt characteristics. Summary of the Invention

[0004] The purpose of this invention is to provide an adaptive water volume control system and method for floor scrubbers to solve the problems mentioned in the background art.

[0005] To address the aforementioned technical problems, one objective of this invention is to provide a floor scrubber water volume adaptive adjustment control system, comprising a dirt identification module, a dual-channel collaborative decision-making module, and a dual-channel control execution module, wherein: The dirt identification module is used to identify the type and extent of dirt on the floor to be cleaned. The dual-channel collaborative decision-making module is used to receive the dirt type and dirt range output by the dirt identification module, and map the dirt type to N sets of water output intensity values ​​and the dirt range to M sets of water output volume values ​​according to the preset mapping relationship, forming N×M combination scenarios, determining the combination scenario corresponding to the best cleaning effect, and outputting the water output intensity value and water output volume value. The dual-channel control execution module is used to control the floor scrubber to perform water discharge operation based on the water discharge intensity and water discharge volume values ​​output by the dual-channel collaborative decision module.

[0006] Preferably, the dirt identification module includes a pre-identification unit, a post-verification unit, and a verification comparison unit, wherein: The front-end recognition unit is positioned at the very front of the floor scrubber's travel direction, before the roller brush installation location. It simultaneously acquires reflectance spectrum images of the ground using multiple spectral bands, and fuses these multi-band spectral images using an information entropy weighted fusion method to output a fused image. Based on a preset target detection model, it detects and locates stains in the fused image, obtaining the stain's coordinate information and boundary contour, thus determining the initial recognition result of the dirt range. Based on a preset target classification model, it classifies the stains in the corresponding areas of the multi-band spectral images, identifying the initial recognition result of the dirt type, including but not limited to oily dirt, watery dirt, adhesive solid residue, particulate dust, and mixed dirt. The post-verification unit is located behind the roller brush and installed inside the wastewater recovery pipe. After the roller brush passes over the surface to be cleaned and completes the cleaning, the wastewater containing dirt flows into the wastewater tank through the recovery pipe. The turbidity value and color temperature value of the wastewater are detected. Based on the turbidity and color temperature values, the characteristics of the actual cleaned dirt are deduced, and the verification result is output, wherein: When the turbidity value is higher than the first turbidity threshold and the color temperature value is lower than the first color temperature threshold, it is determined to be grease-like dirt; When the turbidity value is between the second turbidity threshold and the third turbidity threshold and the color temperature value is higher than the second color temperature threshold, it is determined to be mineral-based or detergent residue-based dirt. The verification adhesion coefficient is determined based on the reverse dirt characteristics, and the verification adhesion coefficient is compared with the initial adhesion coefficient output by the front-end recognition unit (110). If the difference between the verification adhesion coefficient and the initial adhesion coefficient is greater than the preset deviation threshold, it is determined that there is a deviation between the verification result and the initial recognition result, and forced re-recognition is triggered.

[0007] Preferably, the front-end identification unit further includes a spatial marker reset unit, which is used to establish a spatial mapping table of the identified dirt range based on the current position coordinates of the floor scrubber and the boundary coordinates of the identified dirt range. During the movement of the floor scrubber, the spatial coordinates of the current water spray coverage area are obtained in real time, and the spatial coordinates of the current water spray coverage area are compared with the marked areas in the spatial mapping table. If there is an overlap, the dirt identification signal of the overlapping area is ignored.

[0008] Preferably, the dual-channel collaborative decision-making module includes a force mapping unit, a water volume mapping unit, and a combined search unit; The force mapping unit has a built-in first preset mapping relationship, which maps N different types of dirt to N different water output force values, where N is an integer greater than or equal to 2. The dirt type represents the stubbornness of the dirt. The more stubborn the dirt, the greater the mapped water output force value. The water volume mapping unit has a built-in second preset mapping relationship, which maps M different dirt range levels to M different water output values, where M is an integer greater than or equal to 2. The dirt range represents the amount of dirt per unit area. The larger the dirt range, the larger the mapped water output value. The combined lookup unit is used to combine the first preset mapping relationship and the second preset mapping relationship to form an N×M-dimensional combined scene lookup table, so that each entry in the combined scene lookup table corresponds to a cleaning scene formed by a combination of a specific dirt type and a specific dirt range, and outputs the optimal water output force value and the optimal water output volume value according to the experimental calibration library that has been pre-tested and calibrated to achieve the best cleaning effect.

[0009] Preferably, the experimental calibration library includes: For each cleaning scenario formed by a combination of specific dirt types and specific dirt ranges, a gridded scanning experiment is conducted within the two-dimensional parameter space of water output force and water output value with a preset step size. In each experiment, the floor scrubber was controlled to perform standard cleaning operations on the sample to be cleaned with the current parameter combination. After cleaning, at least several cleaning effect indicators, including dirt removal rate, water stain residue on the ground, comprehensive energy consumption and cleaning time, were measured. Substitute the aforementioned multiple cleaning effect indicators into a preset multi-objective comprehensive evaluation function to calculate a comprehensive score; The parameter combination that maximizes the comprehensive score within the parameter space is determined as the optimal water output strength and optimal water output value for this cleaning scenario.

[0010] Preferably, the dual-channel control execution module includes a first control channel and a second control channel. The first control channel is a water outlet force control channel, including a roller brush motor driver and a roller brush motor, which controls the water outlet force by adjusting the output power of the roller brush motor. The second control channel is a water outlet volume control channel, including a water supply valve driver and a water supply valve, which controls the water outlet volume by adjusting the opening duration and opening size of the water supply valve. The first control channel and the second control channel are independent execution channels.

[0011] Preferably, the dual-channel control execution module further includes a real-time feedback adjustment module, used to continuously acquire cleaning effect parameters during the cleaning process. These parameters include the current change rate of the roller brush motor, the turbidity change rate in the wastewater recovery pipe, and the detection value of residual water stains on the ground. Based on the changing trends of these cleaning effect parameters, the module dynamically adjusts the water output intensity and flow rate parameters, wherein: When the current change rate is detected to be continuously increasing while the turbidity change rate is continuously decreasing, it is determined that the current water output force is insufficient and the water output force is increased. The PWM duty cycle of the roller brush motor is increased to increase the mechanical input energy. When the detected value of residual water stains on the ground is higher than the preset water stain threshold, it is determined that the current water output is too large and the water output is reduced. The opening duration of the water supply valve is reduced to decrease the water supply per unit area. When the turbidity change rate drops below the preset cleaning completion threshold, it is determined that the current dirt area has been cleaned, the enhanced cleaning mode of that area is terminated in advance and switched to energy-saving mode.

[0012] Preferably, the dual-channel control execution module further includes a scene control unit, which is used to set a first threshold and a second threshold, including the following combined scenes: The first combination scenario corresponds to small areas of stubborn dirt with a stubbornness level higher than the first threshold and a dirt range lower than the second threshold. The output is a high water output force value higher than the first force threshold and a low water output value lower than the first water volume threshold. The second combination scenario corresponds to a large area of ​​light dirt with a stubbornness level lower than the third threshold and a dirt range higher than the fourth threshold. The output is a low water output strength value lower than the second strength threshold and a high water output value higher than the second water volume threshold.

[0013] The second objective of this invention is to provide a method for adaptive water volume adjustment and control of a floor scrubber, including the floor scrubber adaptive water volume adjustment and control system described in any one of the above-mentioned methods, comprising the following steps: The dirt recognition module identifies the type and extent of dirt on the floor to be cleaned. The dual-channel collaborative decision-making module receives the dirt type and dirt range output by the dirt identification module, maps the dirt type to N sets of water output intensity values ​​and the dirt range to M sets of water output volume values ​​according to a preset mapping relationship, forming N×M combination scenarios, and determines the combination scenario that achieves the best cleaning effect, and outputs the water output intensity value and water output volume value. The dual-channel control execution module controls the floor scrubber to perform water output operations based on the water output intensity and flow rate values ​​output by the dual-channel collaborative decision-making module. The real-time feedback adjustment module continuously acquires cleaning effect parameters during the cleaning process. These cleaning effect parameters include the current change rate of the roller brush motor, the turbidity change rate in the wastewater recovery pipe, and the detection value of residual water stains on the ground. The water output intensity and flow rate parameters are dynamically adjusted based on the changing trends of these cleaning effect parameters.

[0014] Compared with the prior art, the beneficial effects of the present invention are as follows: On the one hand, this invention identifies the type and extent of dirt on the floor to be cleaned through a dirt identification module, overcoming the poor reliability of single-sensor identification. It achieves simultaneous perception and output of dual-dimensional information on dirt type and extent, avoiding the hardware costs of redundant sensors. On the other hand, the dual-channel collaborative decision-making module decouples dirt type and extent into two independent control dimensions and outputs optimal parameters based on an N×M-dimensional combined scenario experimental calibration library. This achieves near-optimal cleaning efficiency and water and energy conservation, outputting high power only for stubborn dirt and actively reducing speed for light dirt, significantly extending motor life. Simultaneously, it ensures that the water supply is just enough to fully wet the dirt without forming a barrier water film, maximizing the cleaning contribution of each milliliter of water. This allows the intensity and water volume to be independently optimized to their respective optimal values ​​for floor washing process control, avoiding the dilemma of forced waste to improve efficiency or forced sacrifice of effect to save resources in traditional coupled control.

[0015] Moreover, this invention continuously monitors the rate of change of the roller brush motor current, the rate of change of sewage turbidity, and the detection value of residual water stains on the ground during the cleaning process through a real-time feedback adjustment module. When the rate of change of current continues to rise while the rate of change of turbidity continues to fall, the water output intensity is automatically increased. When the residual water stains exceed the threshold, the water output is automatically reduced. When the rate of change of turbidity drops below the cleaning completion threshold, the enhanced mode is terminated immediately and the system switches to the energy-saving mode. This achieves dynamic termination based on cleaning results rather than fixed-duration cleaning based on time. This not only effectively avoids cleaning bottlenecks caused by dry friction and water stains caused by excessive water supply, but also maximizes the saving of water and electricity resources and reduces mechanical wear while ensuring cleaning effect. At the same time, the real-time feedback adjustment module and the post-verification unit form a clear division of labor between in-process adjustment and post-process verification, which helps to maintain near-optimal cleaning performance when facing uncalibrated working conditions, significantly improving the system's adaptability and intelligence level. Attached Figure Description

[0016] Figure 1 This is a schematic diagram of the overall structure of Example 1; Figure 2 This is a block diagram illustrating the principle of the real-time feedback adjustment module in Example 1. The meanings of the labels in the diagram are as follows: 100. Dirt identification module; 200. Dual-channel collaborative decision-making module; 300. Dual-channel control execution module. Detailed Implementation

[0017] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0018] Example 1 like Figure 1 As shown, one of the objectives of this invention is to provide a floor scrubber water volume adaptive adjustment control system, including a dirt recognition module 100, a dual-channel collaborative decision-making module 200, and a dual-channel control execution module 300. The dirt recognition module 100 is used to identify the type and extent of dirt on the floor to be cleaned; Specifically, the dirt recognition module 100 includes a pre-recognition unit, a post-verification unit, and a verification and comparison unit, wherein: The front-mounted recognition unit is positioned at the very front of the floor scrubber's travel direction, approximately 5-10cm before the roller brush. This ensures that image acquisition of the target area is completed before the roller brush contacts the ground and water is sprayed. Specifically, a multispectral camera is used, capable of simultaneously acquiring the ground's reflectance spectrum image across multiple spectral bands (covering visible light 400-700nm, near-infrared 700-1100nm, and short-wave infrared 1100-2500nm). Dirt with different chemical compositions exhibits characteristic spectral responses in specific bands; for example, grease-based dirt shows characteristic spectral responses in the near-infrared band (approximately 930nm and 1200nm). The wavelength (00nm) exhibits overtone absorption characteristics of CH bonds, while protein-based dirt shows characteristic absorption peaks of NH bonds in the short-wave infrared band. An information entropy-weighted fusion method is used to fuse multi-band spectral images. Different spectral bands contain different amounts of information; bands with high information entropy contain richer details, while bands with low information entropy may contain more noise or redundant information. The information entropy of each band is calculated, and bands with information entropy higher than a preset entropy threshold are given higher fusion weights, making the dirt area more prominent in the fused image. This provides high-quality input images for subsequent target detection, and the output is a fused image. Based on a pre-defined target detection model (such as the YOLO architecture), stains in the fused image are detected and located to obtain the coordinate information and boundary contour of the stains, thereby determining the initial identification result of the range of dirt. Based on a pre-defined target classification model (such as the MobileViT architecture), stains in corresponding regions of multi-band spectral images are classified, and the initial identification results of the dirt type are identified, including but not limited to oily dirt, watery dirt, adhesive solid residues, particulate dust and mixed dirt. Therefore, on the one hand, non-contact optical detection means that the sensor does not come into direct contact with sewage, and there is no performance degradation problem caused by sensor wear or sewage pollution. Its long-term reliability is better than that of contact sewage sensor solutions. On the other hand, it achieves one-time sensing and dual output, avoiding the hardware cost of setting up redundant sensors to obtain data in two dimensions. The post-verification unit is located behind the roller brush and installed inside the wastewater recovery pipe. After the roller brush passes over the surface to be cleaned and completes the cleaning, the wastewater containing dirt flows into the wastewater tank through the recovery pipe. The unit detects the turbidity value (NTU, representing suspended solids concentration) and the color temperature value (K value, representing color characteristics) of the wastewater. Based on the turbidity and color temperature values, the characteristics of the actual cleaned dirt are inferred. High turbidity + low color temperature usually indicates grease-like dirt, while medium turbidity + high color temperature usually indicates mineral or detergent residue. The verification result is output, including: When the turbidity value is higher than the first turbidity threshold and the color temperature value is lower than the first color temperature threshold, it is judged to be oily dirt (high concentration of suspended matter + yellowish temperature). When the turbidity value is between the second and third turbidity thresholds and the color temperature value is higher than the second color temperature threshold, it is determined to be mineral or detergent residue dirt (medium concentration of suspended matter + bluish temperature). The verification adhesion coefficient is determined based on the reverse-engineered dirt characteristics, and the verification adhesion coefficient is compared with the initial adhesion coefficient output by the pre-identification unit (110). If the difference between the verification adhesion coefficient and the initial adhesion coefficient is greater than the preset deviation threshold (recommended range 0.10-0.20), it is determined that there is a deviation between the verification result and the initial identification result, triggering forced re-identification. This facilitates the subsequent suspension of the floor scrubber's forward movement, increases the output power of the roller brush motor and / or the water output, and strengthens the cleaning of the same target area. Then, the identification data of the target area (including pre-dry identification and post-wet verification) is collected again until the identification result is stable or the number of re-identifications reaches the preset upper limit (e.g., 3 times). If the deviation still exists after the number of re-identifications reaches the upper limit, an alarm signal is output to prompt the user that there may be special dirt that is difficult to identify. Otherwise, the initial identification result is confirmed to be valid, preventing misjudgment from being transmitted to the execution level and improving the robustness of the entire control system.

[0019] It is worth noting that the front-end recognition unit also includes a spatial marker reset unit, which is used to establish a spatial mapping table of the identified dirt range based on the current position coordinates of the floor scrubber and the boundary coordinates of the identified dirt range. During the movement of the floor scrubber, the spatial coordinates of the current water spray coverage area are obtained in real time, and the spatial coordinates of the current water spray coverage area are compared with the marked areas in the spatial mapping table. If there is an overlap, the dirt recognition signal of the overlapping area is ignored to prevent excessive water output caused by repeated recognition.

[0020] Secondly, the dual-channel collaborative decision-making module 200 is used to receive the dirt type and dirt range output by the dirt identification module 100, and to map the dirt type to N sets of water output intensity values ​​and the dirt range to M sets of water output volume values ​​according to the preset mapping relationship, forming N×M combination scenarios, determining the combination scenario corresponding to the best cleaning effect, and outputting the water output intensity value and water output volume value. Furthermore, the dual-channel collaborative decision-making module 200 includes a force mapping unit, a water volume mapping unit, and a combined search unit; The power mapping unit has a built-in first preset mapping relationship, which maps N different types of dirt to N different water output power values, where N is an integer greater than or equal to 2. The dirt type represents the stubbornness of the dirt. The more stubborn the dirt, the greater the mapped water output power value. This realizes the differentiated adhesion mechanism of various dirt such as grease (van der Waals forces + capillary adsorption), sugars (crystallization mechanical interlocking), proteins (hydrogen bond cross-linking), and dust (electrostatic adsorption). The distillation is packaged into the first preset mapping relationship of "type → power", which is equivalent to a one-dimensional mapping table. This allows the complex interdisciplinary cleaning chemistry knowledge to be transformed into numbers that the microcontroller can directly look up. High power is only output when encountering stubborn dirt, while the rotation speed is actively reduced in the case of large area of ​​light dirt to improve service life. If there is too little water, the wetting will be insufficient; if there is too much water, a free water film will form, blocking the direct mechanical contact between the roller brush and the ground, which will reduce the wiping efficiency and cause water stains. Therefore, the water volume mapping unit has a built-in second preset mapping relationship. The second preset mapping relationship maps M different dirt range levels to M different water output values, where M is an integer greater than or equal to 2. The dirt range represents the amount of dirt per unit area. The larger the dirt range, the larger the mapped water output value. This is conducive to ensuring that the dirt per unit area is fully wetted and forms a sewage mixture that can be recovered under negative pressure. This allows each milliliter of water to make the greatest cleaning contribution and achieves the ultimate efficiency of water resource utilization. The combined lookup unit is used to construct an N×M-dimensional combined scenario lookup table by combining the first preset mapping relationship and the second preset mapping relationship. Each entry in the combined scenario lookup table corresponds to a cleaning scenario formed by a combination of a specific dirt type and a specific dirt range, ensuring the completeness of full-condition coverage. Based on the stored experimental calibration library that has been pre-calibrated to achieve the best cleaning effect, it outputs the optimal water output force value and the optimal water output volume value. The force is optimized with cleaning efficiency as the main optimization goal, and the water volume is optimized with resource consumption and water stain control as the main optimization goals. The decombined lookup unit allows both to be independently optimized to their respective optimal values, avoiding the dilemma in traditional coupled control where waste is forced to be accepted in order to improve efficiency or effect is forced to be sacrificed in order to save resources. For the first time, the cleaning efficiency index and the water-saving and energy-saving index are simultaneously approached to the optimal in the same system.

[0021] Specifically, the experimental calibration library includes: For each cleaning scenario formed by a combination of specific dirt types and specific dirt ranges, a gridded scanning experiment is conducted within the two-dimensional parameter space of water output force and water output value with a preset step size. In each experiment, the floor scrubber was controlled to perform standard cleaning operations on the sample to be cleaned with the current parameter combination. After cleaning, multiple cleaning effect indicators were measured, including at least the dirt removal rate, the amount of water residue on the ground, the overall energy consumption, and the cleaning time. The dirt removal rate is the percentage of dirt removed after the cleaning operation relative to the initial total dirt. Using a calibrated spectrophotometer, nine measurement points were evenly selected in the cleaning area to measure the amount of dirt before and after cleaning. The dirt removal rate is calculated as (amount of dirt before cleaning - amount of dirt after cleaning) / amount of dirt before cleaning. The amount of water residue on the ground is the average thickness of the water film remaining on the ground per unit area after the cleaning operation is completed and a standard waiting time (5 seconds) has elapsed. An infrared reflective moisture sensor array was used to scan point by point at a grid density of 1cm × 1cm to obtain the water film at each measurement point. Thickness: The arithmetic mean of the water film thickness at all measurement points is taken as the residual water stain value on the ground. Comprehensive energy consumption is the weighted comprehensive value of electrical energy and water consumed to complete the cleaning of a unit area (1m²). A high-precision DC power meter is used to connect a current sampling resistor in series at the power input terminal of the roller brush motor driver, and a voltage measurement is connected in parallel. The additional energy consumption caused by the increase in cleaning intensity is calculated as the comprehensive energy consumption, that is, the basic operating power consumption of the floor scrubber (basic power consumption of the travel motor, suction motor, etc.) is deducted. Cleaning time is the effective cleaning time required to complete the cleaning of a unit area (1m²). The cleaning start time is the moment when the roller brush motor current shows a significant increase (ΔI>threshold), and the cleaning completion time is the moment when the sewage turbidity change rate drops below the preset threshold. Cleaning time = (cleaning completion time - cleaning start time) / cleaning area. Multiple cleaning performance indicators are substituted into a preset multi-objective comprehensive evaluation function to calculate the comprehensive score. The multi-objective comprehensive evaluation function is S = W1×C - W2×R - W3×E - W4×T, where C is the dirt removal rate, W1 is the weight of the dirt removal rate (0.4), R is the amount of water stains remaining on the ground (0.25), W2 is the weight of the amount of water stains remaining on the ground (0.25), E is the comprehensive energy consumption (0.2), W3 is the weight of the comprehensive energy consumption (0.2), T is the cleaning time (0.15), and W4 is the weight of the cleaning time (0.15). The weights can be verified and fine-tuned through experimental methods. The combination of parameters that maximizes the overall score within the parameter space is determined as the optimal water output strength and optimal water output value for this cleaning scenario.

[0022] Thirdly, the dual-channel control execution module 300 is used to control the floor scrubber to perform water discharge operation based on the water discharge force value and water discharge volume value output by the dual-channel collaborative decision module 200; Specifically, the dual-channel control execution module 300 includes a first control channel and a second control channel. The first control channel is a water output force control channel, which includes a roller brush motor driver and a roller brush motor. The water output force is controlled by adjusting the output power of the roller brush motor. The second control channel is a water output volume control channel, which includes a water supply valve driver and a water supply valve. The water output volume is controlled by adjusting the opening duration and opening size of the water supply valve. The first control channel and the second control channel are independent execution channels.

[0023] It is worth noting that, in order to dynamically fine-tune the parameters based on the actual response signals during the cleaning process, so that the system can still maintain near-optimal performance in uncalibrated scenarios, such as... Figure 2As shown, the dual-channel control execution module (300) also includes a real-time feedback adjustment module, which is used to continuously acquire cleaning effect parameters during the cleaning process. The cleaning effect parameters include the current change rate of the roller brush motor, the turbidity change rate in the sewage recovery pipe, and the detection value of residual water stains on the ground. Based on the changing trend of the cleaning effect parameters, the water output force parameters and water output volume parameters are dynamically adjusted. Specifically, the current change rate of the roller brush motor reflects the magnitude of the mechanical power output of the motor through the working current in the roller brush motor drive circuit. When the friction between the roller brush and the ground increases (i.e., the cleaning force demand increases), the mechanical load of the motor increases, and the drive current rises accordingly. The rate of change of current, ΔI / Δt, is the speed at which the current changes per unit time, reflecting the changing trend of cleaning resistance. The turbidity change rate within the wastewater recovery pipe is achieved using a diffuse light turbidity sensor (such as the TS-100 type, range 0-1000 NTU), installed in the wastewater recovery pipe at the wastewater inlet behind the roller brush (approximately 15-20 cm from the roller brush). This ensures that the collected wastewater is the most recently generated cleaning material, not the mixed liquid in the wastewater tank. During the cleaning process, the change in turbidity reflects the rate at which dirt is stripped from the ground and enters the recovery pipe. The rate of change of turbidity, ΔNTU / Δt, is the speed at which turbidity changes per unit time. Higher cleaning efficiency results in a faster initial rise in turbidity; dirt is... After cleaning, the turbidity tends to stabilize and the rate of change approaches zero. The residual water stain detection value is obtained by installing an infrared reflective moisture sensor array at the rear of the floor scrubber chassis (about 10cm behind the roller brush and in front of the squeegee). A total of 16 sensors in a 4×4 grid pattern (1cm spacing between adjacent sensors) are used to cover a 4cm×4cm area. The value represents the average thickness (unit: μm) of the residual water film on the ground after cleaning. The thicker the water film, the higher the risk of the ground being slippery, and the easier it is to form water stains after the water evaporates. Water has a strong absorption characteristic for infrared light of a certain wavelength (such as 940nm). The thicker the residual water film on the ground, the greater the proportion of infrared light absorbed, and the weaker the light intensity reflected back to the sensor. in: When the current change rate is detected to be continuously increasing (positive values ​​for 3 consecutive sampling windows) while the turbidity change rate is continuously decreasing, it is determined that the current water output force is insufficient and the water output force is increased. The PWM duty cycle (force) of the roller brush motor is increased to increase the mechanical input energy. Insufficient water output force causes dry friction between the roller brush and the dirt, and the cleaning enters a bottleneck state. The adhesion bottleneck is broken through by increasing the speed and torque. The adjustment strategy adopts an incremental PID method: each adjustment is 5% of the current value, and the maximum does not exceed the upper limit value calibrated in the experiment (e.g., PWM≤90%) to avoid excessive deviation from the optimal range. When the detected value of residual water stains on the ground is higher than the preset water stain threshold (50μm in this embodiment), it is determined that the current water output is too large and the water output is reduced. The opening duration (or opening size) of the water supply valve is reduced to reduce the water supply per unit area. The reduction strategy is: based on the current water volume value, reduce it by 10% each time, and the minimum value shall not be lower than the experimentally calibrated lower limit value (e.g., valve opening time ≥80ms) to prevent insufficient wetting due to insufficient water volume. When the turbidity change rate drops below the preset cleaning completion threshold (0.5 NTU / s in this embodiment), it is determined that the current dirt area has been cleaned. The enhanced cleaning mode of this area is terminated in advance and switched to the energy-saving mode. The enhanced cleaning mode of the current area is immediately exited (e.g., the PWM is reduced from 80% to 20% and the water volume is reduced from 150ms to 50ms), and the low-power maintenance mode is switched to. Only the basic brushing and drying functions are maintained. When the next dirt area is reached, the identification and enhanced cleaning process is triggered again. This achieves result-oriented cleaning termination rather than time-oriented fixed-duration cleaning, avoiding incomplete cleaning due to premature termination and waste of water and electricity due to late termination, thus saving water and electricity consumption and reducing unnecessary mechanical wear. Furthermore, the real-time feedback adjustment module and the front-end identification unit form a composite control architecture of feedforward + feedback. The front-end identification provides initial parameter feedforward based on the type and range of dirt, and the real-time feedback adjustment performs closed-loop correction based on the actual cleaning response. The combination of the two can still maintain near-optimal cleaning performance when facing uncalibrated working conditions. At the same time, the real-time feedback adjustment module and the post-verification unit (120) form a division of labor between in-process adjustment and post-process verification. In-process real-time correction ensures the efficiency of the execution process, and post-process verification ensures the acceptability of the final result. Together, they form a complete control closed loop from pre-judgment → in-process correction → post-process confirmation.

[0024] Furthermore, to avoid blindly increasing the water volume and compromising the dry scraping effect, and to avoid blindly increasing the force and wasting energy, the dual-channel control execution module 300 also includes a scene control unit. The scene control unit is used to set a first threshold and a second threshold as constraint boundaries of the strategy to prevent deviation from the experimentally calibrated optimal value, including the following combined scenarios: The first combination scenario corresponds to small areas of stubborn dirt with a stubbornness level higher than the first threshold and a dirt range lower than the second threshold. It outputs a high water output force value higher than the first force threshold and a low water output value lower than the first water volume threshold, realizing a "high concentration, small water volume, high force" cleaning mode for small areas of stubborn dirt. A small amount of water fully wets the dirt to soften it, and high mechanical force breaks off the adhesion, which not only ensures the cleaning effect, but also avoids a large amount of water forming water stains in a small area. The second scenario corresponds to large areas of light dirt with a stubbornness level below the third threshold and a dirt range above the fourth threshold. It outputs a low water pressure value below the second pressure threshold and a high water volume value above the second water volume threshold, realizing a "large area, low pressure, medium water volume" cleaning mode for large areas of light dirt. Sufficient water evenly covers a large area to suspend loose dirt, which can be efficiently removed with gentle brushing, avoiding unnecessary energy consumption and wear caused by high pressure.

[0025] The second objective of this invention is to provide a floor scrubber water volume adaptive adjustment control system, applicable to any of the above-mentioned floor scrubber water volume adaptive adjustment control methods, comprising the following steps: The dirt identification module 100 identifies the type and extent of dirt on the floor to be cleaned. The dual-channel collaborative decision-making module 200 receives the dirt type and dirt range output by the dirt identification module 100, maps the dirt type to N sets of water output intensity values ​​and the dirt range to M sets of water output volume values ​​according to the preset mapping relationship, forming N×M combination scenarios, and determines the combination scenario corresponding to the best cleaning effect, and outputs the water output intensity value and water output volume value. The dual-channel control execution module 300 controls the floor scrubber to perform water output operation based on the water output intensity and water output values ​​output by the dual-channel collaborative decision-making module 200. The real-time feedback adjustment module continuously acquires cleaning effect parameters during the cleaning process, including the current change rate of the roller brush motor, the turbidity change rate in the wastewater recovery pipe, and the detection value of residual water stains on the ground. Based on the changing trend of the cleaning effect parameters, the water output intensity and water output parameters are dynamically adjusted.

[0026] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely preferred examples and are not intended to limit the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the present invention as claimed. The scope of protection of the present invention is defined by the appended claims and their equivalents.

Claims

1. A floor scrubber water volume adaptive adjustment control system, characterized in that: It includes a dirt recognition module (100), a dual-channel collaborative decision-making module (200), and a dual-channel control execution module (300), wherein: The dirt identification module (100) is used to identify the type and extent of dirt on the surface to be cleaned; The dual-channel collaborative decision-making module (200) is used to receive the dirt type and dirt range output by the dirt identification module (100), and map the dirt type to N sets of water output intensity values ​​and the dirt range to M sets of water output volume values ​​according to the preset mapping relationship, forming N×M combination scenarios, measuring the combination scenario corresponding to the best cleaning effect, and outputting the water output intensity value and water output volume value. The dual-channel control execution module (300) is used to control the floor scrubber to perform water discharge operation based on the water discharge force value and water discharge volume value output by the dual-channel collaborative decision module.

2. The adaptive water volume adjustment control system for floor scrubbers according to claim 1, characterized in that: The dirt identification module (100) includes a pre-identification unit, a post-verification unit, and a verification comparison unit, wherein: The front-end recognition unit is positioned at the very front of the floor scrubber's travel direction, before the roller brush installation location. It simultaneously acquires reflectance spectrum images of the ground using multiple spectral bands, and fuses these multi-band spectral images using an information entropy weighted fusion method to output a fused image. Based on a preset target detection model, it detects and locates stains in the fused image, obtaining the stain's coordinate information and boundary contour, thus determining the initial recognition result of the dirt range. Based on a preset target classification model, it classifies the stains in the corresponding areas of the multi-band spectral images, identifying the initial recognition result of the dirt type, including but not limited to oily dirt, watery dirt, adhesive solid residue, particulate dust, and mixed dirt. The post-verification unit is located behind the roller brush and installed inside the wastewater recovery pipe. After the roller brush passes over the surface to be cleaned and completes the cleaning, the wastewater containing dirt flows into the wastewater tank through the recovery pipe. The turbidity value and color temperature value of the wastewater are detected. Based on the turbidity and color temperature values, the characteristics of the actual cleaned dirt are deduced, and the verification result is output, wherein: When the turbidity value is higher than the first turbidity threshold and the color temperature value is lower than the first color temperature threshold, it is determined to be grease-like dirt; When the turbidity value is between the second turbidity threshold and the third turbidity threshold and the color temperature value is higher than the second color temperature threshold, it is determined to be mineral-based or detergent residue-based dirt. The verification adhesion coefficient is determined based on the reverse dirt characteristics, and the verification adhesion coefficient is compared with the initial adhesion coefficient output by the front-end recognition unit (110). If the difference between the verification adhesion coefficient and the initial adhesion coefficient is greater than the preset deviation threshold, it is determined that there is a deviation between the verification result and the initial recognition result, and forced re-recognition is triggered.

3. The adaptive water volume adjustment control system for floor scrubbers according to claim 2, characterized in that: The front-end identification unit also includes a spatial marker reset unit, which is used to establish a spatial mapping table of the identified dirt range based on the current position coordinates of the floor scrubber and the boundary coordinates of the identified dirt range. During the movement of the floor scrubber, the spatial coordinates of the current water spray coverage area are obtained in real time, and the spatial coordinates of the current water spray coverage area are compared with the marked areas in the spatial mapping table. If there is an overlap, the dirt identification signal of the overlapping area is ignored.

4. The adaptive water volume adjustment control system for floor scrubbers according to claim 2, characterized in that: The dual-channel collaborative decision-making module (200) includes a force mapping unit, a water volume mapping unit, and a combined search unit; The force mapping unit has a built-in first preset mapping relationship, which maps N different types of dirt to N different water output force values, where N is an integer greater than or equal to 2. The dirt type represents the stubbornness of the dirt. The more stubborn the dirt, the greater the mapped water output force value. The water volume mapping unit has a built-in second preset mapping relationship, which maps M different dirt range levels to M different water output values, where M is an integer greater than or equal to 2. The dirt range represents the amount of dirt per unit area. The larger the dirt range, the larger the mapped water output value. The combined lookup unit is used to combine the first preset mapping relationship and the second preset mapping relationship to form an N×M-dimensional combined scene lookup table, so that each entry in the combined scene lookup table corresponds to a cleaning scene formed by a combination of a specific dirt type and a specific dirt range, and outputs the optimal water output force value and the optimal water output volume value according to the experimental calibration library that has been pre-tested and calibrated to achieve the best cleaning effect.

5. The floor scrubber water volume adaptive adjustment control system according to claim 4, characterized in that: The experimental calibration library includes: For each cleaning scenario formed by a combination of specific dirt types and specific dirt ranges, a gridded scanning experiment is conducted within the two-dimensional parameter space of water output force and water output value with a preset step size. In each experiment, the floor scrubber was controlled to perform standard cleaning operations on the sample to be cleaned with the current parameter combination. After cleaning, at least several cleaning effect indicators, including dirt removal rate, water stain residue on the ground, comprehensive energy consumption and cleaning time, were measured. Substitute the aforementioned multiple cleaning effect indicators into a preset multi-objective comprehensive evaluation function to calculate a comprehensive score; The parameter combination that maximizes the comprehensive score within the parameter space is determined as the optimal water output strength and optimal water output value for this cleaning scenario.

6. The floor scrubber water volume adaptive adjustment control system according to claim 5, characterized in that: The dual-channel control execution module (300) includes a first control channel and a second control channel. The first control channel is a water output intensity control channel, including a roller brush motor driver and a roller brush motor, which controls the water output intensity by adjusting the output power of the roller brush motor. The second control channel is a water output volume control channel, including a water supply valve driver and a water supply valve, which controls the water output volume by adjusting the opening duration and opening size of the water supply valve. The first control channel and the second control channel are independent execution channels.

7. The adaptive water volume adjustment control system for a floor scrubber according to claim 6, characterized in that: The dual-channel control execution module (300) also includes a real-time feedback adjustment module, used to continuously acquire cleaning effect parameters during the cleaning process. These parameters include the current change rate of the roller brush motor, the turbidity change rate in the wastewater recovery pipe, and the detection value of residual water stains on the ground. Based on the changing trends of these cleaning effect parameters, the module dynamically adjusts the water output intensity and flow rate parameters, wherein: When the current change rate is detected to be continuously increasing while the turbidity change rate is continuously decreasing, it is determined that the current water output force is insufficient and the water output force is increased. The PWM duty cycle of the roller brush motor is increased to increase the mechanical input energy. When the detected value of residual water stains on the ground is higher than the preset water stain threshold, it is determined that the current water output is too large and the water output is reduced. The opening duration of the water supply valve is reduced to decrease the water supply per unit area. When the turbidity change rate drops below the preset cleaning completion threshold, it is determined that the current dirt area has been cleaned, the enhanced cleaning mode of that area is terminated in advance, and the energy-saving mode is switched.

8. The floor scrubber water volume adaptive adjustment control system according to claim 7, characterized in that: The dual-channel control execution module (300) also includes a scene control unit, which is used to set a first threshold and a second threshold, including the following combined scenes: The first combination scenario corresponds to small areas of stubborn dirt with a stubbornness level higher than the first threshold and a dirt range lower than the second threshold. The output is a high water output force value higher than the first force threshold and a low water output value lower than the first water volume threshold. The second combination scenario corresponds to a large area of ​​light dirt with a stubbornness level lower than the third threshold and a dirt range higher than the fourth threshold. The output is a low water output strength value lower than the second strength threshold and a high water output value higher than the second water volume threshold.

9. A method for adaptive water volume adjustment and control of a floor scrubber, applied to the adaptive water volume adjustment and control system of the floor scrubber as described in any one of claims 1-8, characterized in that, Includes the following steps: The dirt identification module (100) identifies the type and extent of dirt on the surface to be cleaned; The dual-channel collaborative decision-making module (200) receives the dirt type and dirt range output by the dirt identification module (100), maps the dirt type to N sets of water output intensity values ​​and the dirt range to M sets of water output volume values ​​according to the preset mapping relationship, forming N×M combination scenarios, and determines the combination scenario corresponding to the best cleaning effect, and outputs the water output intensity value and water output volume value. The dual-channel control execution module (300) controls the floor scrubber to perform water output operation based on the water output intensity and water output value output by the dual-channel collaborative decision module (200). The real-time feedback adjustment module continuously acquires cleaning effect parameters during the cleaning process. The cleaning effect parameters include the current change rate of the roller brush motor, the turbidity change rate in the sewage recovery pipe, and the detection value of residual water stains on the ground. The water output intensity and water output parameters are dynamically adjusted according to the changing trend of the cleaning effect parameters.

Images