A control method, device, medium and program product of a mobile robot in a cross-temperature zone environment

Abstract

A control method and device of a mobile robot in a cross-temperature zone environment, a medium and a program product, relate to the technical fields of intelligent warehouse logistics and robot automatic control. The control device quantifies the complex multi-parameter environment state (temperature, humidity) into a risk index by calculating the condensation safety margin. On this basis, the sensing parameters are adjusted from two directions of expanding the shielding radius to filter near-field noise points and reducing the threshold to tolerate signal attenuation. When the risk is removed, the parameter reset instruction timely restores the full performance operation state. Therefore, the navigation continuity and operation stability of the mobile robot in the cross-temperature zone condensation working condition are realized, and the task interruption caused by environmental mutation is avoided.

Description

Technical Field

[0001] This application relates to the fields of intelligent warehousing and logistics and robot automatic control technology, and in particular to a control method, device, medium and program product for a mobile robot in a cross-temperature environment. Background Technology

[0002] With the rapid development of intelligent manufacturing and automated logistics technologies, autonomous mobile robots (AMRs) have become key equipment for internal logistics in factories.

[0003] In related technologies, to address potential optical interference issues for robot sensors in low-temperature environments (cold storage) (such as the rapid condensation that can occur on sensor surfaces when a robot quickly moves from a low-temperature environment to a high-temperature, high-humidity environment), cryogenic AMRs typically incorporate resistance heating wires inside or around the protective housing of the LiDAR or vision camera, or are equipped with dedicated hot air blowers. An independent temperature control circuit controls the heating element to operate continuously or activate upon detecting a low temperature. This utilizes external electrical energy to directly heat the lens or glass cover on the sensor surface, preventing ice formation or removing frost, thus ensuring the sensor's continuous operation in low-temperature environments.

[0004] However, in standard industrial logistics scenarios, factories may deploy standard AMR charging stations in corridors connecting workshops, semi-outdoor awnings, or non-temperature-controlled auxiliary areas to optimize space utilization and meet fire safety regulations, while the robots' main tasks are concentrated in temperature- and humidity-controlled production workshops. This necessitates frequent switching between low-temperature charging and parking areas and high-temperature, high-humidity production areas during task execution. Furthermore, the internal structure of standard AMRs differs from that of low-temperature AMRs, typically lacking pre-reserved electrical interfaces and installation space for high-power defrosting and heating devices. Adding hardware to existing fleets is not only costly but may also compromise sensor sealing and existing warranty structures. Environmental changes leading to condensation can also force robots to stop at boundary areas to wait for the condensation to dissipate, reducing the continuity and efficiency of logistics handling. Summary of the Invention

[0005] This application provides a control method, device, medium, and program product for mobile robots operating in cross-temperature environments, which can improve the operational continuity of standard mobile robots when switching between cross-temperature environments.

[0006] In a first aspect, this application provides a control method for a mobile robot operating in a cross-temperature zone environment, applied to a control device. The method includes: when the control device determines that the target mobile robot has ended its charging state after a preset time and has received a target task, it determines the current body temperature data of the target mobile robot and the current environmental data of the target operating area corresponding to the target task; the control device calculates a dew point temperature value based on the current environmental data and a saturated vapor pressure formula, and uses the difference between the current body temperature data and the dew point temperature value as a condensation safety margin; when the control device determines that the condensation safety margin is less than a preset first safety threshold, it determines a point cloud shielding radius based on the condensation safety margin and a pre-stored noise spatial distribution model, wherein the noise spatial distribution model characterizes the positive correlation between the radial distribution standard deviation of condensation noise points on the surface of an optical sensor and the condensation risk value; the control device, based on... The condensation safety margin and the preset light transmission attenuation function are used to calculate the intensity compensation coefficient. Based on the intensity compensation coefficient, the control device lowers and corrects the preset standard reflection intensity threshold to obtain the current execution threshold. The current execution threshold is used by the target mobile robot to identify reflector features in the navigation path. After establishing a point cloud filtering envelope with the point cloud shielding radius as the boundary, the control device generates a navigation parameter configuration instruction that includes the point cloud filtering envelope and the current execution threshold, and sends the navigation parameter configuration instruction to the target mobile robot. After determining that the target mobile robot responds to the navigation parameter configuration instruction and moves towards the target work area, and if the condensation safety margin rises to a level greater than the preset second safety threshold, the control device generates a parameter reset instruction and sends it to the target mobile robot. The parameter reset instruction configuration includes restoring the navigation parameters to their default values.

[0007] By employing the aforementioned technical solution, the control equipment quantifies complex multi-parameter environmental conditions (temperature, humidity) into risk indicators by calculating the condensation safety margin. Based on this, through the synergistic effect of a noise spatial distribution model and a light transmission attenuation function, the system simultaneously adjusts the sensing parameters from two directions: expanding the shielding radius to filter near-field noise and reducing the threshold tolerance signal attenuation. This dual compensation mechanism suppresses frequent sudden stops caused by false obstacles generated by condensation droplets and prevents positioning failures caused by the loss of the true reflector due to signal weakening. Once the risk is eliminated, the parameter reset command promptly restores the robot to full-performance operation. Therefore, the mobile robot achieves navigation continuity and operational stability under cross-temperature condensation conditions, avoiding task interruptions caused by sudden environmental changes.

[0008] In conjunction with some embodiments of the first aspect, in some embodiments, the step of determining the point cloud shielding radius based on the condensation safety margin and a pre-stored noise spatial distribution model, wherein the noise spatial distribution model characterizes the positive correlation between the radial distribution standard deviation of condensation noise points on the optical sensor surface and the condensation risk value, specifically includes: mapping the condensation safety margin to the pre-stored noise spatial distribution model to obtain a noise boundary value covering a pre-set signal interval, wherein the noise spatial distribution model includes the condensation safety margin value and the condensation noise dispersion being negatively correlated; and determining the noise boundary value as the point cloud shielding radius.

[0009] By employing the above technical solution, when mapping the condensation safety margin to the noise spatial distribution model, the model explicitly defines a negative correlation between the safety margin value and the condensation noise dispersion (i.e., the lower the margin, the higher the risk and the wider the noise distribution). Based on this relationship, it outputs a noise boundary value covering a pre-set information interval (e.g., 99.7% of the noise range). This solution can dynamically adjust the shielding boundary according to the real-time condensation level, avoiding both excessively small shielding ranges leading to residual noise causing misjudgments and excessively large shielding ranges causing collision risks due to the mis-shielding of real near-field obstacles, thus achieving a balance between noise suppression and effective sensing range.

[0010] In conjunction with some embodiments of the first aspect, in some embodiments, the step of generating a navigation parameter configuration instruction that includes the point cloud filtering envelope and the current execution threshold, and sending the navigation parameter configuration instruction to the target mobile robot, specifically includes: after the control device determines the preliminary navigation path of the target mobile robot based on the target task and a preset global navigation map, it identifies road segments in the preliminary navigation path where the temperature gradient change rate exceeds a preset threshold, and marks the road segments as condensation risk areas; when the control device determines that the distance between the positioning coordinates of the target mobile robot and the boundary of the condensation risk area is less than a preset trigger threshold, it sends the navigation parameter configuration instruction to the target mobile robot.

[0011] By adopting the above technical solution, the control equipment proactively identifies and marks condensation risk areas by analyzing the rate of change of temperature gradient in the initial navigation path, and only issues degradation parameter configuration commands when the robot approaches the boundary of the area (the distance is less than a preset trigger threshold). This geofencing-based precise triggering mechanism avoids premature activation of restrictive parameters, which would cause the robot to operate at low performance for extended periods in normal areas. It achieves a dynamic balance between safety protection and operational efficiency, improving the overall time efficiency of cross-temperature zone logistics operations.

[0012] In conjunction with some embodiments of the first aspect, in some embodiments, before the step of the control device generating a navigation parameter configuration instruction containing the point cloud filtering envelope surface bounded by the point cloud shielding radius, and sending the navigation parameter configuration instruction to the target mobile robot, the method further includes: the control device establishing a kinematic braking safety inequality containing the point cloud shielding radius, maximum deceleration, and system response delay time based on the braking performance parameters and system response delay time of the target mobile robot, wherein the braking performance parameters include the maximum deceleration under wet and slippery road surface conditions in the target working area; the control device solving for the maximum permissible transition speed that satisfies the collision-free constraint condition based on the kinematic braking safety inequality, wherein the maximum permissible transition speed is used to write the navigation parameter configuration instruction and to limit the travel speed of the target mobile robot before the condensation safety margin is restored.

[0013] By adopting the above technical solution, after determining the point cloud shielding radius (i.e., the artificially created perception blind zone), the control device does not adjust the perception parameters in isolation, but simultaneously establishes a kinematic braking safety inequality that includes the shielding radius, the maximum deceleration under slippery road conditions, and the system response delay time. By solving this inequality to obtain the maximum permissible transition speed, the robot's kinetic energy is limited to a safe range that allows it to stop in time even if an obstacle suddenly appears in the blind zone. This establishes a rigid coupling mechanism between perception degradation and motion limitation, eliminating the near-field collision hazard that may be caused by expanding the shielding radius.

[0014] In conjunction with some embodiments of the first aspect, in some embodiments, the control device solves for the maximum permissible transition speed that satisfies the collision-free constraint based on the kinematic braking safety inequality. This maximum permissible transition speed is used to write the navigation parameter configuration instruction and to limit the travel speed of the target mobile robot before the condensation safety margin is restored. Specifically, this includes: the control device constructing a first braking distance function based on the ratio of the square of the speed to twice the maximum deceleration; the control device constructing a second reaction distance function based on the product of the speed and the system response delay time; the control device iteratively solving for the maximum positive real number solution of the speed variable with the constraint that the sum of the first braking distance function and the second reaction distance function is less than the difference between the preset effective detection distance of the sensor and the shielding radius of the point cloud; and the control device determining the maximum positive real number solution as the maximum permissible transition speed.

[0015] By adopting the above technical solution, the control equipment decomposes the complex kinematic braking safety problem into two independent physical models: a braking distance function based on the ratio of the square of velocity to deceleration, and a reaction distance function based on the product of velocity and time delay. By constructing a constraint that the sum of these two functions must be less than the difference between the effective detection distance and the shielding radius, and using an iterative method to solve for the maximum positive real number solution of the velocity variable, this physical modeling-based solution method, compared to empirical fixed speed limit strategies, can adapt to the current shielding radius size, road friction conditions, and system response characteristics. It ensures both the rigid requirements of braking safety and avoids the significant decrease in operational efficiency caused by overly conservative speed limits, achieving an optimal balance between safety and traffic efficiency.

[0016] In conjunction with some embodiments of the first aspect, in some embodiments, after sending the navigation parameter configuration command to the target mobile robot, the method further includes: the control device receiving a set of candidate reflector point clouds fed back by the target mobile robot based on the current execution threshold; the control device constructing a current topological polygon based on the spatial coordinates of each feature point in the candidate reflector point cloud set, and performing geometric feature matching between the current topological polygon and the standard reflector topology in a preset map; if the control device determines that the similarity of the geometric feature matching is lower than a preset topological integrity threshold, it generates an inertial navigation switching command and sends it to the target mobile robot, the inertial navigation switching command being configured to control the target mobile robot to enable dead reckoning for navigation.

[0017] By adopting the above technical solution, the control equipment does not blindly trust the adjusted perception system after issuing degradation parameters. Instead, it constructs the current topological polygon by receiving the candidate reflector point cloud set from the robot and performs geometric feature matching with the standard reflector topology in the preset map. When the matching similarity is lower than the topological integrity threshold, it indicates that even with parameter adjustments, there are still serious reflector loss or misidentification issues. At this point, an inertial navigation switching command is immediately generated to activate dead reckoning for navigation. This avoids complete robot loss of control or on-site alarm shutdown due to the failure of a single navigation source under extreme condensation conditions, ensuring that even under harsh conditions where optical sensors are nearly completely ineffective, basic passage tasks can still be completed using the inertial unit.

[0018] In conjunction with some embodiments of the first aspect, in some embodiments, before the step of determining the point cloud shielding radius based on the condensation safety margin and the pre-stored noise spatial distribution model, the method further includes: when the control device determines that the condensation safety margin is less than a preset first safety threshold, generating a waste heat compensation control command and sending it to the target mobile robot, wherein the waste heat compensation control command is configured to trigger the target mobile robot to perform high-load floating-point operations that do not occupy business logic.

[0019] By adopting the above technical solution, when the control equipment detects that the condensation safety margin is lower than the first safety threshold, it prioritizes generating a waste heat compensation control command to trigger the robot to perform high-load floating-point operations. Although this operation does not participate in the business logic, it uses the power consumption and heat generated by the CPU or GPU chip to transfer waste heat to the surface of the optical sensor through the heat conduction path inside the robot (such as the physical contact between the heat sink and the sensor housing), actively raising the sensor temperature from a physical level to move it away from the dew point. This reduces the degree of restriction on perception and movement, and improves the speed and safety margin for crossing temperature zones.

[0020] In a second aspect, this application provides a control device comprising: one or more processors and a memory; the memory is coupled to the one or more processors and is used to store computer program code including computer instructions, wherein the one or more processors invoke the computer instructions to cause the control device to perform the method described in the first aspect and any possible implementation thereof.

[0021] Thirdly, this application provides a computer program product containing instructions that, when run on a control device, cause the control device to perform the method described in the first aspect and any possible implementation thereof.

[0022] Fourthly, this application provides a computer-readable storage medium including instructions that, when executed on a control device, cause the control device to perform the method described in the first aspect and any possible implementation thereof.

[0023] One or more technical solutions provided in the embodiments of this application have at least the following technical effects or advantages:

[0024] 1. By adopting a dual-parameter adaptive adjustment mechanism based on the dynamic generation of point cloud shielding radius and the degraded reflection intensity threshold according to the condensation safety margin, the technical problem of optical sensor failure due to cross-temperature zone condensation in standard mobile robots under the condition of no hardware defrosting is effectively solved, thus ensuring the continuous operation capability and navigation stability of mobile robots in extreme cross-temperature zone environments.

[0025] 2. Due to the adoption of a multimodal navigation automatic switching mechanism based on geometric topology feature verification (i.e., after reducing the reflection intensity threshold to improve the recall rate, geometric matching is performed by constructing the current topological polygon and the standard reflector structure, and automatically switching to inertial navigation mode when the similarity is insufficient), the technical problem of robot positioning drift caused by forcibly matching incorrect landmarks when extreme condensation causes the lidar to completely fail or the data is extremely discrete is effectively solved in the existing technology. Thus, it realizes the ability to maintain short-term high-precision blind running navigation capability in the optical blind zone by relying on dead reckoning by utilizing the advantages of multi-source sensor fusion.

[0026] 3. By adopting the above technical solution, this control device creatively utilizes the thermodynamic side effects of the robot's onboard computing unit as an active defense mechanism. When the risk of condensation approaches the safety threshold, software instructions trigger the underlying hardware to perform high-load floating-point operations to generate waste heat, and utilize thermal conduction to raise the temperature of the robot body and sensor surfaces. This reduces the temperature difference between the sensors and the ambient dew point, thereby achieving continuous operation. Attached Figure Description

[0027] Figure 1 This is a flowchart illustrating a control method for a mobile robot operating in a temperature-zone-dependent environment, as described in an embodiment of this application.

[0028] Figure 2 This is another flowchart illustrating a control method for a mobile robot operating in a cross-temperature zone environment, as described in this application.

[0029] Figure 3 This is a schematic diagram of an exemplary hardware structure of the control device in an embodiment of this application. Detailed Implementation

[0030] The terminology used in the following embodiments of this application is for the purpose of describing particular embodiments only and is not intended to be limiting of this application. As used in the specification and appended claims of this application, the singular expressions “a,” “an,” “the,” “the,” “the,” and “this” are intended to include the plural expressions as well, unless the context clearly indicates otherwise. It should also be understood that the term “and / or” as used in this application refers to and includes any or all possible combinations of one or more of the listed items.

[0031] Hereinafter, the terms "first" and "second" are used for descriptive purposes only and should not be construed as implying or suggesting relative importance or implicitly indicating the number of indicated technical features. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature, and in the description of the embodiments of this application, unless otherwise stated, "multiple" means two or more.

[0032] To facilitate understanding, the relevant terms and concepts involved in the embodiments of this application will be introduced below.

[0033] Autonomous Mobile Robot (AMR): Refers to an automated transportation device with autonomous navigation capabilities, used to perform material handling tasks in production or warehousing environments. Specifically, a mobile robot refers to an intelligent hardware terminal capable of sensing its environment and planning its path using sensors; it can also refer to industrial vehicles equipped with optical sensing modules such as LiDAR or vision sensors, such as AGVs with holstering and lifting capabilities or forklift-type AMRs. In this application's context, mobile robots specifically refer to automated equipment that needs to frequently move between different temperature zones (such as semi-outdoor charging areas and indoor constant-temperature operating areas).

[0034] Please see Figure 1 This is a flowchart illustrating a control method for a mobile robot operating in a temperature-zone-dependent environment, as described in this application.

[0035] S101. When the control device determines that the target mobile robot will end its charging state after a preset time and receives the target task, it determines the current body temperature data of the target mobile robot and the current environmental data of the target working area corresponding to the target task.

[0036] Among them, the control equipment refers to the host computer system or edge computing node used for centralized scheduling, status monitoring and task distribution of mobile robots. It can be used to represent a server cluster or industrial computer (IPC) with data processing and communication capabilities, and usually runs a robot scheduling system (RCS) that can acquire the robot's status telemetry data and issue control commands. The target task refers to the logistics operation instructions issued by the upper-level manufacturing execution system (MES) or warehouse management system (WMS), which includes information such as the starting position, the ending position, the type of goods and the priority. The target operation area is used to represent the spatial range covered by the target task execution path. The current body temperature data refers to the real-time temperature reading of key components inside the robot (such as the battery pack or the lidar shell).

[0037] This step is typically executed at a specific point in time when the robot is about to leave the charging station, in a scenario where the robot is in a semi-outdoor or low-temperature charging environment and is about to enter an indoor constant-temperature working environment. Specifically, the control device first polls the status of each robot's Battery Management System (BMS). When it detects that a robot's battery level has reached a preset threshold (e.g., 95%) or N minutes (preset duration) before receiving a charging completion signal, it determines that the robot is in a "about to end charging" state. At the same time, the control device checks the task queue. If there are tasks assigned to the robot, it obtains the robot's body temperature transmitted back by its BMS through the established TCP / IP or MQTT communication link with the robot. It also obtains real-time environmental temperature and humidity data through environmental IoT sensors deployed at the workshop entrance and next to the production line, thereby constructing a temperature difference context between the robot's "current state" and the "environment it is about to enter".

[0038] In some embodiments, the control device sends a status query frame to the target mobile robot and parses the temperature field in the returned message. At the same time, based on the path planning information of the target task, the control device identifies the ID of the nearest environmental sensor on the path, requests real-time temperature and relative humidity data from the environmental sensor via an industrial bus (such as Modbus TCP), and synchronizes and associates these two sets of data using timestamps as indexes.

[0039] In some embodiments, when the target mobile robot detects that the charging current has dropped to the trickle charging stage, it actively pushes a "ready" event packet to the control device. This event packet encapsulates the current onboard temperature sensor value. After responding to the event, the control device directly reads the real-time environmental field data of that area from a pre-built digital twin factory model. It is understood that environmental data can also be obtained by using historical data weighted prediction, which is not limited here.

[0040] It is understandable that other redundant data source fusion methods can also be used to obtain the body temperature, such as combining historical temperature curves for Kalman filtering prediction, or using the measurement value of the contact temperature probe on the charging pile side as a backup data source when the BMS temperature sensor fails. No limitation is made here.

[0041] It should be further explained that when the target work area spans multiple temperature and humidity zones, the control equipment should identify all zones along the robot's planned path, query the environmental parameters of each zone, and calculate the weighted average of the path or select the parameters of the worst working condition (maximum temperature difference, highest humidity) as a conservative estimate. For example, if the path needs to pass through a normal temperature zone (20℃) and a high humidity zone (25℃ / 80%RH), the parameters of the high humidity zone should be used to ensure that the risk assessment is not underestimated.

[0042] S102. The control equipment calculates the dew point temperature value based on the current environmental data and the saturated water vapor pressure formula, and uses the difference between the current body temperature data and the dew point temperature value as the condensation safety margin.

[0043] The saturated vapor pressure formula is a thermodynamic formula describing the relationship between the maximum partial pressure of water vapor that air can hold at a specific temperature and temperature. Commonly used approximate expressions include the Magnus-Tetens formula or the Goff-Gratch equation. The dew point temperature is the temperature threshold at which air is cooled to saturation (relative humidity reaching 100%) under constant pressure and water vapor content. When the surface temperature of an object is below this value, water vapor will condense into liquid water on the surface.

[0044] The control device extracts the air temperature from the current environmental data. and relative humidity Calculate the current dew point temperature using the Magnus formula. The typical calculation logic is as follows: ,in For auxiliary functions, It is a constant. After obtaining the dew point temperature, the control equipment reads the robot's body temperature. Perform subtraction: This difference is the "condensation safety margin". If the difference is positive and large, it means that the robot is relatively hot and not easy to condense; if the difference is close to zero or negative, it means that the surface temperature of the robot is lower than the ambient dew point, and water vapor can easily liquefy on the sensor surface to form a condensate film.

[0045] In some embodiments, dew point and margin calculations can be performed in several ways: Optionally, the control device uses the standard Magnus formula for calculation. The control device calls a pre-set mathematical function library, inputs the ambient temperature and relative humidity, directly calculates the dew point temperature, and then subtracts the dew point temperature from the robot body temperature to obtain a scalar form safety margin value. Optionally, the control device uses a lookup table method for rapid estimation. The control device has a pre-stored lookup table for dew point under different temperature and humidity combinations. Based on the interval index of the current environmental data, it directly looks up the corresponding dew point temperature value, introduces a set correction factor (such as considering the thermal conductivity of the sensor housing material) to correct the body temperature, and then performs the difference calculation. It is understood that other simplified formulas can also be used to achieve rapid estimation, such as using empirical formulas. No restrictions are imposed here.

[0046] It should be noted that in extreme low-temperature environments (T_env<0℃), water vapor condenses directly into solid ice crystals rather than liquid water droplets. In this case, the saturated water vapor pressure formula for ice surfaces (such as the Goff-Gratch equation for ice) should be used for correction calculations to avoid underestimating the risk of icing. At the same time, in high-altitude or negative pressure environments, an atmospheric pressure correction factor needs to be introduced because the saturated water vapor pressure is proportional to the air pressure.

[0047] S103. When the control device determines that the condensation safety margin is less than the preset first safety threshold, it determines the point cloud shielding radius based on the condensation safety margin and the pre-stored noise spatial distribution model.

[0048] Among them, the noise spatial distribution model refers to a mathematical model constructed based on the optical characteristics of the sensor and thermodynamic experimental data. It is used to characterize the radial distribution law of artifact noise points in space with the sensor as the origin under different degrees of condensation. The point cloud shielding radius represents the boundary distance of the circular area that can be directly filtered out without feature extraction. It is used to represent the near-field perception blind zone setting value of the robot.

[0049] When the control device detects that the real-time calculated condensation safety margin has fallen below the first safety threshold, it indicates that an interfering water film is about to form or has already formed on the sensor lens surface. The control device calls the locally stored noise spatial distribution model, which can be a pre-trained multivariate regression function or a discretized lookup table data structure. The lower the robot body temperature is compared to the ambient dew point (the smaller the margin, or even negative), the larger and denser the water vapor droplets liquefied on the sensor lens surface become. These droplets have a stronger scattering and refraction effect on the laser beam, resulting in a false echo that is not only high in intensity but also more spatially diffused. Conversely, if the margin is only slightly below the safety threshold, only a small amount of mist forms at the lens edge, and the noise is concentrated in a very close range. Through the mapping calculation of this model, the control device finally outputs a quantized point cloud shielding radius value, which is then encapsulated in the navigation parameter configuration data packet for subsequent transmission and application.

[0050] In some embodiments, based on a continuous mapping method using a statistical model, the control device maps the condensation safety margin to a pre-stored noise spatial distribution model. This model defines a negative correlation between the condensation safety margin value and the condensation noise dispersion (e.g., the standard deviation of a Gaussian distribution) (i.e., the lower the margin, the greater the dispersion). The control device calculates a coverage area within a pre-set confidence interval (e.g., ...) using the model. The noise boundary value (within a certain range) is directly determined as the point cloud shielding radius, thus achieving stepless smooth adjustment of the radius according to the risk level. Optionally, the control device has a preset multi-level risk ladder table, dividing the condensation safety margin below the first safety threshold into "primary condensation zone", "intermediate condensation zone" and "severe condensation zone". The control device determines which interval the current margin falls into, retrieves the corresponding fixed radius value (e.g., corresponding to 0.15 meters, 0.3 meters, and 0.45 meters respectively), and determines the retrieved value as the point cloud shielding radius.

[0051] Understandably, in practical deployments, a hybrid adaptive optimization approach can also be used to dynamically determine the point cloud shielding radius. Specifically, during initial operation, the control device uses the aforementioned discrete lookup table method to quickly provide an initial radius value. Subsequently, during the robot's task execution, the system analyzes the raw laser point cloud data stream transmitted back by the robot in real time, statistically analyzing features such as point cloud density distribution characteristics, point cloud intensity variance, and point cloud temporal stability within and outside the current shielding radius. Online learning algorithms are then used to continuously correct and update the parameters of the noise spatial distribution model, gradually bringing the model's output shielding radius closer to the optimal value under the current real-world conditions. This adaptive mechanism can address model drift issues caused by temperature and humidity changes under different seasons and weather conditions, improving the system's long-term stability and generalization ability. Further limitations are not specified here.

[0052] Specifically, after issuing the initial shielding radius configuration to the robot, the real-time point cloud data stream from the LiDAR transmitted back by the robot via the network is analyzed. The data analysis module of the control device performs layered processing on this point cloud data. First, a subset of the point cloud located near the boundary of the currently set shielding envelope is extracted. For example, if the current shielding radius is R, the extracted subset is located within the distance range of [R-0.05 meters]. All point cloud points within [R+0.1 m] are used as observation samples. Multi-dimensional feature analysis is then performed on the point cloud within this ring area. The first analysis indicator is the point cloud density mutation characteristic. The control device calculates the point cloud number density within a unit solid angle range in this ring area. If the detected density is significantly higher than the background density of point clouds in the normal environment (e.g., more than 3 times higher), and these high-density point clouds exhibit high-frequency inter-frame flicker characteristics (i.e., severe position and intensity jitter in several consecutive frames), it is determined to be a characteristic of condensation noise, indicating that the current shielding radius may be too small, posing a risk of noise leakage. The second analysis indicator is the verification of the geometric continuity and physical rationality of the point cloud. The control device performs clustering and segmentation of the point cloud within the ring area, checking for the existence of point cloud clusters forming stable geometric shapes. For example, if a cluster of point clouds maintains a relatively stable columnar or blocky structure for several consecutive frames, and its reflection intensity conforms to the characteristic range of common obstacles (such as human bodies, shelf legs, and wall corners), then these point clouds are likely reflections of real objects rather than condensation. Noise indicates that the current shielding radius may be too large, causing real obstacles to be mistakenly shielded. Based on the comprehensive judgment results of the above two feature analyses, the control device executes dynamic parameter adjustment logic. If it is determined that there is a risk of noise leakage, the control device immediately generates a parameter update command, increases the point cloud shielding radius by a fixed step (e.g., 0.05 meters) on the current value, and reissues it to the robot. At the same time, the triggering reason and magnitude of this correction are recorded in the adjustment log. If it is determined that there is a risk of over-shielding, the control device reduces the shielding radius by a small step (e.g., 0.03 meters) and continuously verifies the adjustment effect in subsequent frames. In order to prevent the parameters from oscillating repeatedly near the boundary conditions, the control device also introduces a hysteresis interval mechanism. That is, the parameter modification operation is only actually executed when the adjustment conditions are met for N consecutive frames (e.g., N=10, corresponding to about 1 second). Through this closed-loop dynamic correction strategy based on real-time feedback, the system can automatically adapt to the actual situation under different environments and different degrees of condensation.

[0053] In some embodiments, before determining the point cloud shielding radius, when the control device determines that the condensation safety margin is less than a first safety threshold, it first generates a waste heat compensation control command and sends it to the target mobile robot via a wireless network. The payload of this command contains a specific control code to trigger the robot's onboard industrial control computer to execute a predefined high-load floating-point operation task that does not occupy business logic, such as starting an empty loop matrix multiplication thread or performing meaningless FFT fast Fourier transform calculations. These calculation operations will cause the CPU and GPU chips to enter a high-power state, thereby generating a large amount of waste heat. Since the lidar is usually installed in the same sealed cabinet as the industrial control computer, the heat is transferred to the sensor housing through air convection and metal thermal conduction, causing its temperature to passively rise during the robot's movement, thereby actively reducing the temperature difference between the sensor and the ambient dew point and reducing the physical intensity of actual condensation. This waste heat utilization strategy leaves more room for adjustment in subsequent parameter optimization. After the waste heat compensation command is issued, the main control logic unit of the control device calls the locally stored noise spatial distribution model, and the control device substitutes the currently acquired condensation safety margin value into the model as an input variable.

[0054] S104. The control device calculates the intensity compensation coefficient based on the condensation safety margin and the preset optical transmission attenuation function.

[0055] The optical transmission attenuation function represents a physical model describing the attenuation of energy of a laser beam as it penetrates a non-uniform medium (such as a water mist layer or a condensate film) with respect to propagation distance or medium density. This function is usually constructed based on the Beer-Lambert law or Mie scattering theory.

[0056] When the laser beam emitted by the lidar passes through a micron-sized water film on the sensor's photomask surface, Rayleigh and Mie scattering occur. This causes some energy to be scattered and lost before reaching the actual target (such as a reflector or obstacle). Furthermore, the returning light path must penetrate the water film again, resulting in secondary attenuation. The control device retrieves a preset optical transmission attenuation function from memory. This function was established during the system development phase by simulating condensation environments (such as applying water mist of varying thicknesses to the sensor lens in a constant temperature and humidity chamber) and measuring the laser echo intensity attenuation rate under different condensation conditions. The control device reads the real-time condensation safety margin (denoted as...). ), and substitute it into the preset optical transmission attenuation function middle.

[0057] In some embodiments, specific function calculations can be implemented in multiple ways:

[0058] Optionally, a piecewise linear decay function can be used to calculate the control device's preset safety threshold. (For example ) and saturation threshold (For example ), and minimum compensation coefficient (For example The strength compensation coefficient of the control equipment is calculated according to the following formula. :

[0059] ,

[0060] This method uses linear interpolation to make the coefficient decrease linearly with the temperature difference, which requires little computation and has controllable logic.

[0061] Optionally, the Sigmoid nonlinear optical density function can be used for calculation: The control equipment utilizes a variant of the Sigmoid function to simulate the nonlinear process of water mist formation (i.e., a slow decrease in transmittance in the initial stage of fogging, a sharp decrease in the middle stage, and a tendency to stabilize in the later stage). The control equipment uses the formula... Perform calculations, where The preset curve steepness factor, This represents the temperature difference offset at the half-decay position. This method can more smoothly fit the physical condensation process and avoid system control oscillations caused by sudden changes in coefficients. Understandably, a discrete mapping method based on a real-time lookup table can also be used, but this is not limited here.

[0062] During implementation, issues may arise due to individual sensor differences (such as varying lens aging rates) leading to inaccurate calculations of the uniform attenuation function, resulting in compensation coefficients that are either too high or too low. The detailed solution to this problem is to employ a reference calibration mechanism in the control equipment. Specifically, the control equipment calculates the theoretical coefficients... Then, the nearest known high-confidence reflector in the current scan frame (determined based on its location on the historical map) will be selected, and its current measured intensity will be compared with that reflector. Its historical normal intensity Calculate the measured attenuation ratio The control equipment will Ratio with actual measurement Perform weighted fusion (e.g.) Thus, the final intensity compensation coefficient that adapts to the actual state of the current sensor is obtained.

[0063] S105. The control device adjusts the preset standard reflection intensity threshold downward based on the intensity compensation coefficient to obtain the current execution threshold.

[0064] The standard reflection intensity threshold refers to the set of intensity filtering reference values ​​(usually a fixed value or a curve that varies with distance) used by the navigation system to distinguish between ordinary environmental features (such as walls and shelf legs) and highly reflective road signs (reflectors) under standard operating conditions of dryness and no condensation.

[0065] The control device first reads a preset standard reflection intensity threshold from the navigation algorithm's configuration database. This threshold is usually determined during system initialization or debugging based on the reflector model and sensor performance calibration. Subsequently, the control device performs mathematical operations on the intensity compensation coefficient (usually less than 1.0) calculated in S104 and the standard threshold, relaxing the judgment criteria by lowering the threshold value.

[0066] In some embodiments, the threshold reduction correction can be achieved in several ways: Optionally, the control device reads a preset standard reflection intensity threshold. (e.g., 2000), and compare it with the strength compensation coefficient. Perform multiplication to obtain the current execution threshold. After calculation, the control device writes the result (e.g., 1200) to a temporary parameter buffer. Subsequently, the control device pushes this execution threshold to the reflector recognition module of the navigation algorithm via the internal bus or shared memory, replacing its original judgment criteria. Optionally, instead of simple multiplication, the control device matches different correction strategies based on the numerical range of the intensity compensation coefficient. When the value is in the high range (e.g., 0.7 to 1.0), the control device adopts a conservative linear downward adjustment, calculated using the following formula: ,in The preset slight correction amount (e.g., 500); when When the value is in the low range (e.g., 0.4 to 0.7), the control device uses an aggressive nonlinear down-adjustment, calculated using the following formula: An exponential term is introduced to enhance the threshold reduction under low-risk coefficients. This approach enables differentiated response strategies under different degrees of condensation. It is understandable that an adaptive threshold regression method based on historically successfully identified samples could also be used; this is not limited here.

[0067] S106. After the control device establishes a point cloud filtering envelope surface with the point cloud shielding radius as the boundary, it generates a navigation parameter configuration instruction containing the point cloud filtering envelope surface and the current execution threshold, and sends the navigation parameter configuration instruction to the target mobile robot.

[0068] The point cloud filtering envelope surface represents a three-dimensional or two-dimensional virtual geometric boundary constructed based on the point cloud shielding radius. All laser point cloud data located inside this boundary will be directly removed from the navigation stack and will not participate in feature matching calculations.

[0069] After completing the shielding radius calculation in S103 and the threshold correction in S105, the control device establishes a circular (or spherical, depending on whether the LiDAR is 2D or 3D) point cloud filtering envelope in the robot's local coordinate system, with the center coordinates of the target mobile robot as the origin and the point cloud shielding radius calculated in S103 as the geometric boundary. The mathematical expression of this envelope can be defined as: all points satisfying... The point cloud data will be discarded directly. Subsequently, the control device encapsulates a navigation parameter configuration command message, typically in JSON or Protocol Buffers format, containing fields such as the mask radius value and the current execution threshold. Mask enable flag, parameter effective timestamp, etc.

[0070] In some embodiments, the generation and transmission of navigation parameter configuration instructions can be implemented in several ways: Optionally, based on the target task and a preset global navigation map, the control device first calls a path planning algorithm to determine the initial navigation path of the target mobile robot. Along each node of this path, the control device queries the environmental temperature attribute data bound to the map (this temperature data can be obtained from historical records, fixed sensors, or thermodynamic simulation predictions), and calculates the rate of change of temperature gradient between adjacent nodes. When the rate of change of gradient of a certain path segment exceeds a preset threshold, the control device marks that segment as a "condensation risk area" at the logic layer and records the boundary coordinates of the area in the map database. Subsequently, the control device monitors the positioning coordinates returned by the target mobile robot through odometry or SLAM algorithm in real time, and calculates the Euclidean distance between these coordinates and the nearest boundary of the condensation risk area. Only when determined (When a preset trigger threshold, such as 5 meters, is reached, the control device sends navigation parameter configuration commands to the target mobile robot via a wireless network, enabling precise regional parameter switching.) Optionally, the robot's onboard environmental sensors can provide real-time feedback data. The control device subscribes to the robot's environmental data stream, and when it detects a drastic fluctuation in the real-time ambient temperature transmitted by the robot within a short time window (such as a rapid temperature rise), it will detect this fluctuation. When the robot is crossing a temperature zone boundary, the control equipment determines that it is doing so and immediately generates and sends navigation parameter configuration instructions. This method does not require pre-marked maps and is highly adaptable, but it requires a high response speed from the sensors.

[0071] Understandably, a triggering method based on virtual geofencing can also be used, setting virtual boundaries at key locations such as the workshop entrance. Once the robot enters the boundary, the trigger parameters are sent out. This is not limited here.

[0072] S107. After the control device determines that the target mobile robot responds to the navigation parameter configuration command and moves to the target work area, and determines that the condensation safety margin has risen to a level greater than the preset second safety threshold, it generates a parameter reset command and sends it to the target mobile robot.

[0073] Once the robot successfully traverses the high-risk condensation area and enters the indoor constant-temperature operating area using degraded parameters, the temperature of the sensor surface will gradually rise due to continuous environmental heat exchange or waste heat compensation, and the condensed water vapor will naturally evaporate and dissipate. The control equipment continuously monitors the robot's status. When the calculated condensation safety margin (body temperature - dew point temperature) continues to rise and stably exceeds the preset second safety threshold, the control equipment determines that the physical risk has been eliminated. At this point, in order to restore the robot's full-function perception capabilities (such as eliminating near-field blind spots and restoring the ability to recognize distant weak reflectors), the control equipment generates a parameter reset command and sends it to the robot, instructing it to unload the point cloud filter envelope and revert to the standard intensity threshold.

[0074] In some embodiments, the control device requests the robot to upload a current laser point cloud quality statistical report. The control device analyzes two indicators in the report: "Near-field noise density" and "Average reflector intensity." Only when the near-field noise density drops to near zero and the reflector intensity recovers to a normal level (e.g., above 90% of the standard value), the control device ignores temperature readings and directly sends a reset command based on perceived facts. It is understood that the robot's absolute positioning area (e.g., having penetrated 50 meters into the core indoor area) can also be used as a forced reset condition; this is not limited here.

[0075] In some embodiments, the control device employs a graded, progressive reset strategy, generating a multi-level reset command sequence. For example, it first restores the intensity threshold (allowing the robot to see clearly into the distance), observes for 10 seconds to confirm that the localization variance is normal, and then gradually reduces the point cloud shielding radius (e.g., from 0.5m -> 0.2m -> 0m) to gradually release the near-field sensing area. If a decrease in localization quality or a surge in noise is detected at any stage, the control device immediately reverts to the previous level of parameters until the sensor state is completely stable, thereby achieving a smooth and imperceptible mode switch.

[0076] In this embodiment, by adopting a point cloud shielding radius and intensity compensation coefficient dynamically generated based on condensation safety margin, and by transforming physical temperature difference risk into a quantifiable sensing parameter adjustment strategy, the navigation failure problem caused by transient condensation on sensor surface in cross-temperature zone environments of standard mobile robots in the prior art is effectively solved, thereby achieving the technical effect of ensuring the robot's safe passage through condensation risk areas without hardware modification.

[0077] In the above embodiments, the control device can achieve adaptive adjustment of navigation parameters through point cloud shielding and intensity compensation. However, in practical applications, when the above method is executed, there is a risk that the effective detection distance will be shortened due to the expansion of the near-field blind zone, which may lead to insufficient braking safety margin.

[0078] Please see Figure 2 This is another flowchart illustrating a control method for a mobile robot operating in a temperature-zone-dependent environment, as described in this application.

[0079] S201. When the control device determines that the target mobile robot will end its charging state after a preset time and receives the target task, it determines the current body temperature data of the target mobile robot and the current environmental data of the target working area corresponding to the target task.

[0080] S202. The control equipment calculates the dew point temperature value based on the current environmental data and the saturated water vapor pressure formula, and uses the difference between the current body temperature data and the dew point temperature value as the condensation safety margin.

[0081] S203. When the control equipment determines that the condensation safety margin is less than the preset first safety threshold, it determines the point cloud shielding radius based on the condensation safety margin and the pre-stored noise spatial distribution model.

[0082] S204. The control equipment calculates the intensity compensation coefficient based on the condensation safety margin and the preset optical transmission attenuation function.

[0083] S205. The control device adjusts the preset standard reflection intensity threshold downward based on the intensity compensation coefficient to obtain the current execution threshold.

[0084] Steps S201~S205 and Figure 1 The steps S101 to S105 in the illustrated embodiment are similar, and can be referred to the descriptions in steps S101 to S105, which will not be repeated here.

[0085] S206. Based on the braking performance parameters of the target mobile robot and the system response delay time, the control equipment establishes a kinematic braking safety inequality that includes the point cloud shielding radius, maximum deceleration, and system response delay time.

[0086] The control device first reads braking performance parameters from the robot's hardware configuration database. These parameters are typically determined experimentally by the robot manufacturer and stored in the vehicle controller's EEPROM. The control device extracts the current load status (empty or fully loaded) and queries the corresponding nominal maximum deceleration value. Simultaneously, through statistical analysis of historical operating data, the control device calculates the average response latency under the current hardware and software architecture. This latency includes the difference between the hardware timestamp of sensor data acquisition from the FPGA, the software timestamp of speed commands generated after processing by the ROS navigation stack, and the communication timestamp of commands sent from the CAN bus to the motor driver. Subsequently, the control device establishes a kinematic braking safety inequality from the detection of an obstacle at the sensor boundary (i.e., the blind zone edge) to the vehicle's complete braking, ensuring that the total displacement is less than or equal to the blind zone radius. This ensures that even in the worst-case scenario (where the obstacle is exactly at the blind zone boundary), the robot will not collide with the obstacle.

[0087] In some embodiments, the inequality can be established in several ways: Optionally, the control device adopts a conservative constraint modeling method based on the effective line-of-sight of the sensor. The control device first obtains the measured effective detection range of the lidar under the current ambient humidity and particulate matter concentration (this distance is dynamically determined by analyzing the signal-to-noise ratio curve of the point cloud in real time, rather than using the theoretical maximum range on the device nameplate), and records it as the dynamic line-of-sight parameter. The control device then reads the currently effective point cloud shielding radius from S103 and calculates the difference between the two to obtain the net usable detection range. Based on this, the control device introduces the product of the system response delay time and the speed as the first constraint (inertial blind driving distance), and then introduces the square of the speed divided by twice the maximum deceleration as the second constraint (pure braking distance). It is required that the sum of the two terms multiplied by the safety margin coefficient (usually taken as 1.2 to 1.5) must still be strictly less than the net usable detection range, thus forming a quadratic inequality constraint on speed. This inequality implicitly transforms the geometric loss of perception degradation into a numerical restriction on the degrees of freedom of motion.

[0088] Optionally, the control device employs a dynamic constraint modeling approach based on accumulated risk probabilities. Instead of directly using a fixed shielding radius value, the control device treats the shielding radius as a probability distribution region with fuzzy boundaries (because the edges of the condensed water mist have gradual characteristics and are not sharply defined). The control device generates multiple virtual obstacle position samples near the shielding radius using a Monte Carlo sampling method, calculates the collision probability at the current speed for each sample, and seeks the expected value of the collision probabilities for all samples. The control device requires this expected value to be lower than a preset safety threshold (e.g., collision probability < 0.001%). By iteratively scanning different speed values, it finds the speed critical point that satisfies the probability constraints and uses this critical point as an equivalent expression of the inequality constraints. It is understood that a rolling time-domain optimization method based on model predictive control (MPC) can also be used to establish the dynamic inequality; this is not limited here.

[0089] S207. The control equipment solves for the maximum allowable transition speed that satisfies the collision-free constraint condition based on the kinematic braking safety inequality.

[0090] The control device treats the kinematic braking safety inequality in S206 as a nonlinear equation with respect to the current velocity. Since the inequality may contain higher-order terms (such as cubic terms considering Jerk's equations) or environmental correction parameters, the control device typically uses numerical analysis to find the maximum value within the domain that makes the inequality true.

[0091] The control device initiates numerical solving and then determines whether the inequality has a real domain of solutions. If the discriminant is negative, it indicates that a non-zero speed that meets safety requirements cannot be found under the current parameter combination, and the system must force the robot to stop and issue an alarm. If a domain of solutions exists, the control device determines the theoretical upper limit of the speed by finding the critical boundary value of the inequality (i.e., transforming the inequality into an equation). Since it is necessary to avoid the robot running at the critical speed for a long time in practical engineering applications (where the fault tolerance margin is zero), the control device introduces a conservative coefficient (such as 0.85) on top of the theoretical upper limit and issues it to the robot as the final maximum permissible speed in transition, ensuring that the system always operates within a range with sufficient safety margin.

[0092] In some embodiments, speed can be solved and optimized in several ways: Optionally, the control device employs an analytical-based exact solution method. The control device expands the inequality constructed in S206 into a standard quadratic inequality with respect to speed, extracts the coefficients of the quadratic, linear, and constant terms, and calls the built-in algebraic root-finding function library (such as using the Cardano formula or the numerically stable Citardauq formula) to calculate the zeros and solution domain boundaries of the inequality. The control device filters out all positive real solutions (negative speeds and imaginary solutions are physically meaningless), selects the maximum value as the upper bound of the theoretically permissible speed, and then multiplies this upper bound by a preset safety reduction factor (this factor is calibrated based on historical accident statistics, typically ranging from 0.75 to 0.9) to obtain the maximum permissible speed value for engineering applications, and writes this value into the robot's speed planner configuration parameters. Optionally, the control device employs an iterative search-based robust solution method. Instead of directly solving the inequality, the control device uses a binary search algorithm to iteratively verify the speed within a preset speed range (e.g., from 0 to the robot's physical limit speed). The control device first selects the midpoint of the range as a candidate speed, substitutes this speed value into the inequality S206, and compares the values ​​on both sides. If the inequality holds, the lower bound of the range is moved up to the midpoint; if it does not hold, the upper bound is moved down to the midpoint. This process is repeated iteratively until the range length converges to a preset precision (e.g., 0.01 m / s). The maximum value within this range is the allowable speed that satisfies the safety constraints. It is understandable that gradient descent-based optimization algorithms can also be used to solve the constrained maximum speed problem; this is not a limitation here.

[0093] S208. After the control device establishes a point cloud filtering envelope with the point cloud shielding radius as the boundary, it generates a navigation parameter configuration instruction containing the point cloud filtering envelope and the current execution threshold, and sends the navigation parameter configuration instruction to the target mobile robot.

[0094] After completing the parameter derivation in steps S203 to S207, the control device encapsulates these scattered decision results and reliably transmits them to the field robot. The control device first instantiates a virtual geometric envelope data structure in memory. This data structure stores the boundary definition of the envelope in the form of an array or vector. For 2D LiDAR, it is typically defined as the radius parameter of a circular region; for 3D LiDAR, it may be defined as a height range plus a cylindrical parameter with a circular base. The control device then packages the geometric description parameters of the envelope, the current execution intensity threshold calculated in S205, and the maximum permissible transition speed obtained in S207 into a standardized message body. This message body follows a predefined communication protocol format (such as a custom JSON structure or an industry-standard OPC-UA object model). The control device tracks the robot's coordinate position in real time. When the straight-line distance between the robot and the nearest risk area boundary decreases to a preset trigger threshold (this threshold must be greater than the robot's braking distance to ensure sufficient time for parameter switching), the control device sends the encapsulated navigation parameter configuration command to the robot's onboard controller via an industrial Ethernet or 5G wireless link.

[0095] S209. After the control device determines that the target mobile robot has responded to the navigation parameter configuration command and moved to the target work area, and determines that the condensation safety margin has risen to a level greater than the preset second safety threshold, it generates a parameter reset command and sends it to the target mobile robot.

[0096] After the robot safely enters the room through the condensation zone in a low-speed, high-filtration "defense mode," the condensate on the sensor surface gradually evaporates as the physical environment changes (temperature increases, humidity decreases). The control equipment continuously monitors the condensation safety margin (the changes at this point are mainly driven by the increase in the robot's body temperature). When this margin not only exceeds the risk threshold but also stably exceeds the preset second safety threshold (introducing a hysteresis range to prevent oscillation), the control equipment determines that the physical risk has been completely eliminated. At this point, to restore the robot's productivity, the control equipment generates a parameter reset command.

[0097] In this embodiment, by employing the collaborative technical features of fusion sensing parameter degradation and kinematic constraints, a kinematic braking safety inequality is established that includes the point cloud shielding radius, system response delay, and braking performance. Based on this inequality, the process of deriving the maximum allowable speed during transition is solved in reverse. This achieves the continuity and safety of mobile robots operating across temperature zones with zero hardware modification cost.

[0098] The following describes an exemplary control device 300 provided in an embodiment of this application. Figure 3 This is a schematic diagram of an exemplary hardware structure of the control device 300 provided in an embodiment of this application.

[0099] In some embodiments, the control device 300 is a computer device or includes a computer device. The computer device includes a processor, memory, and a network interface connected via a system bus. The processor of the computer device provides computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and internal memory. The non-volatile storage medium stores an operating system, computer programs, and a database. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage medium. The database of the computer device stores data. The network interface of the computer device is used to communicate with other external terminals or servers via a network connection. In some embodiments, the network interface can be a wired network interface; in some embodiments, the network interface can also be a wireless network interface. When the computer program is executed by the processor, it implements the methods in the embodiments of this application.

[0100] Those skilled in the art will understand that Figure 3 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0101] The above-described embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit it. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of this application.

[0102] As used in the above embodiments, depending on the context, the term "when..." can be interpreted as meaning "if...", "after...", "in response to determining...", or "in response to detecting...". Similarly, depending on the context, the phrase "when determining..." or "if (the stated condition or event) is interpreted as meaning "if determining...", "in response to determining...", "when (the stated condition or event) is detected", or "in response to detecting (the stated condition or event)".

[0103] In the above embodiments, implementation can be achieved entirely or partially through software, hardware, firmware, or any combination thereof. When implemented using software, it can be implemented entirely or partially in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, all or part of the processes or functions described in the embodiments of this application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., coaxial cable, fiber optic, digital subscriber line) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that integrates one or more available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium (e.g., solid-state drive), etc.

[0104] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. This program can be stored in a computer-readable storage medium, and when executed, it can include the processes described in the above method embodiments. The aforementioned storage medium includes various media capable of storing program code, such as ROM or random access memory (RAM), magnetic disks, or optical disks.

Claims

1. A control method for a mobile robot operating in environments with varying temperature ranges, characterized in that, Applied to control equipment, the method includes: When the control device determines that the target mobile robot will end its charging state after a preset time and receives the target task, it determines the current body temperature data of the target mobile robot and the current environmental data of the target working area corresponding to the target task. The control device calculates the dew point temperature value based on the current environmental data and the saturated water vapor pressure formula, and uses the difference between the current body temperature data and the dew point temperature value as the condensation safety margin. When the control device determines that the condensation safety margin is less than a preset first safety threshold, it determines the point cloud shielding radius based on the condensation safety margin and a pre-stored noise spatial distribution model. The noise spatial distribution model characterizes the positive correlation between the radial distribution standard deviation of condensation noise points on the surface of the optical sensor and the condensation risk value. The control device calculates the intensity compensation coefficient based on the condensation safety margin and the preset optical transmission attenuation function; The control device adjusts the preset standard reflection intensity threshold downward based on the intensity compensation coefficient to obtain the current execution threshold. The current execution threshold is the reflector feature in the navigation path that the target mobile robot identifies based on the current execution threshold. After establishing a point cloud filtering envelope with the point cloud shielding radius as the boundary, the control device generates a navigation parameter configuration instruction containing the point cloud filtering envelope and the current execution threshold, and sends the navigation parameter configuration instruction to the target mobile robot. After the control device determines that the target mobile robot responds to the navigation parameter configuration command and moves toward the target work area, and determines that the condensation safety margin has risen to a level greater than a preset second safety threshold, it generates a parameter reset command and sends it to the target mobile robot. The parameter reset command configuration includes restoring the navigation parameters to their default values.

2. The method according to claim 1, characterized in that, The step of determining the point cloud shielding radius based on the condensation safety margin and a pre-stored noise spatial distribution model, wherein the noise spatial distribution model characterizes the positive correlation between the radial distribution standard deviation of condensation noise points on the optical sensor surface and the condensation risk value, specifically includes: The condensation safety margin is mapped to a pre-stored noise spatial distribution model to obtain a noise boundary value covering a pre-set signal interval. The noise spatial distribution model includes the condensation safety margin value and the condensation noise dispersion negatively correlated. The noise boundary value is determined as the point cloud shielding radius.

3. The method according to claim 1, characterized in that, The step of generating navigation parameter configuration instructions containing the point cloud filtering envelope and the current execution threshold, and sending the navigation parameter configuration instructions to the target mobile robot, specifically includes: After determining the initial navigation path of the target mobile robot based on the target task and the preset global navigation map, the control device identifies road segments in the initial navigation path where the temperature gradient change rate exceeds a preset threshold and marks the road segments as condensation risk areas. When the control device determines that the distance between the positioning coordinates of the target mobile robot and the boundary of the condensation risk area is less than a preset trigger threshold, it sends the navigation parameter configuration command to the target mobile robot.

4. The method according to claim 1, characterized in that, Before the step of the control device establishing a point cloud filtering envelope surface with the point cloud shielding radius as its boundary, generating a navigation parameter configuration instruction containing the point cloud filtering envelope surface and the current execution threshold, and sending the navigation parameter configuration instruction to the target mobile robot, the method further includes: The control device establishes a kinematic braking safety inequality that includes the point cloud shielding radius, maximum deceleration, and system response delay time based on the braking performance parameters of the target mobile robot and the system response delay time. The braking performance parameters include the maximum deceleration under wet and slippery road conditions in the target working area. The control device solves for the maximum permissible transition speed that satisfies the collision-free constraint based on the kinematic braking safety inequality. The maximum permissible transition speed is used to write the navigation parameter configuration instruction and to limit the travel speed of the target mobile robot before the condensation safety margin is restored.

5. The method according to claim 4, characterized in that, The control device solves for the maximum permissible transition speed that satisfies the collision-free constraint based on the kinematic braking safety inequality. This maximum permissible transition speed is used to write the navigation parameter configuration command and to limit the travel speed of the target mobile robot before the condensation safety margin is restored. Specifically, this includes: The control device constructs a first braking distance function based on the ratio of the square of the velocity to twice the maximum deceleration; The control device constructs a second reaction distance function based on the product of speed and system response delay time; The control device uses the constraint that the sum of the first braking distance function and the second reaction distance function is less than the difference between the preset effective detection distance of the sensor and the shielding radius of the point cloud, and iteratively solves for the maximum positive real number solution of the velocity variable. The control device determines the maximum positive real number solution as the maximum allowable transition speed.

6. The method according to claim 1, characterized in that, After the step of sending the navigation parameter configuration command to the target mobile robot, the method further includes: The control device receives a set of candidate reflector point clouds fed back by the target mobile robot based on the current execution threshold; The control device constructs a current topological polygon based on the spatial coordinates of each feature point in the candidate reflector point cloud set, and performs geometric feature matching between the current topological polygon and the standard reflector topological structure in the preset map. When the control device determines that the similarity of the geometric feature matching is lower than a preset topological integrity threshold, it generates an inertial navigation switching command and sends it to the target mobile robot. The inertial navigation switching command is configured to control the target mobile robot to use dead reckoning for navigation.

7. The method according to claim 1, characterized in that, Before the step of determining the point cloud shielding radius based on the condensation safety margin and the pre-stored noise spatial distribution model, the method further includes: When the control device determines that the condensation safety margin is less than a preset first safety threshold, it generates a waste heat compensation control command and sends it to the target mobile robot. The waste heat compensation control command is configured to trigger the target mobile robot to perform high-load floating-point operations that do not occupy business logic.

8. A control device, characterized in that, The control device includes: one or more processors and a memory; the memory is coupled to the one or more processors, the memory is used to store computer program code, the computer program code including computer instructions, and the one or more processors invoke the computer instructions to cause the control device to perform the method as described in any one of claims 1-7.

9. A computer program product containing instructions, characterized in that, When the computer program product is run on a control device, the control device performs the method as described in any one of claims 1-7.

10. A computer-readable storage medium comprising instructions, characterized in that, When the instructions are executed on the control device, the control device performs the method as described in any one of claims 1-7.

Images