Extreme-environment-oriented bulk cargo unloading robot autonomous collaborative operation method and system

Abstract

The invention discloses an extreme environment-oriented bulk cargo unloading robot autonomous collaborative operation method and system. According to the method, Uyuan muon detection units are arranged in an unloading area and on a robot body, the three-dimensional density of a bulk cargo pile and an arch bridge probability voxel field are obtained, cross-modal registration fusion is carried out in combination with charge level geometry and motion data collected by an arm-loaded or vehicle-mounted FMCW millimeter wave radar, and bulk cargo internal state tensors F (x, y, z, t) and a gradient field F thereof are constructed. And the system generates a multi-robot partition and operation sequence according to the real-time calculation result of the unloading potential field U, and automatic unloading, arch breaking and conveying operation are achieved. And when the dust, charge or gas concentration exceeds the limit, the low-energy cooperative mode is automatically switched, the state tensor and the potential field are dynamically updated through the operation receipt, and efficient and safe unloading cooperative control under the extreme environment is achieved.

Description

Technical Field

[0001] This invention relates to the field of intelligent robot control and bulk material unloading technology, specifically to an autonomous collaborative operation method and system for bulk cargo unloading robots in extreme environments. Background Technology

[0002] Existing bulk cargo unloading systems primarily rely on vision, lidar, or ultrasonic methods to detect the surface morphology of the material, and use one or more robots to perform point-to-point material handling and arch breaking operations. However, in extreme environments, such as the unloading of bulk materials like ore, coal, or grain, complex conditions such as high dust, high humidity, highly corrosive aerosols, and low visibility often occur, leading to severe signal attenuation of optical or ultrasonic sensors, data distortion, or even failure, making it difficult to accurately identify internal arches or voids. Furthermore, existing multi-robot systems largely rely on local visual navigation and pre-defined task assignments, lacking a holistic understanding of the dynamic state of the bulk cargo's interior, and are unable to achieve safe and efficient collaborative unloading under conditions of material collapse due to its own weight, changes in flow patterns, or adhesion to the walls.

[0003] Currently, there is no technology capable of real-time 3D state field reconstruction and collaborative control within non-transparent, heavily obscured bulk cargo stacks, nor is there a low-energy sensing fusion method suitable for explosive dust and high electrostatic environments. Therefore, how to establish a visual representation of the interior of bulk cargo stacks under extreme environments, and then realize adaptive zoning and unloading potential field control of multiple robots under safe and robust conditions, has become a key technical problem that urgently needs to be solved in this field. Summary of the Invention

[0004] To address the aforementioned problems, this invention provides an autonomous collaborative operation method and system for bulk cargo unloading robots in extreme environments, which can effectively overcome the shortcomings of existing technologies.

[0005] This invention is achieved through the following technical solution: an autonomous collaborative operation method for bulk cargo unloading robots in extreme environments, comprising: A. Cosmic muon detection units are deployed on the perimeter of the unloading area and on the robot body to obtain multi-view muon scattering / transmission counting data of the bulk cargo pile. Based on the spatiotemporal sliding window, normalization processing and background drift compensation are performed to reconstruct a three-dimensional voxel field representing the density, void and arch probability of the pile. B. The material surface geometry, local flow velocity and collapse boundary are simultaneously acquired by arm-mounted or vehicle-mounted FMCW millimeter-wave radar to obtain near-surface motion observations; C. By fusing the voxel field from step A with the near-surface observations from step B in the same coordinate system through cross-modal registration, the internal state tensor F(x,y,z,t) of the bulk cargo and its gradient field for unloading decisions are constructed. F; D. According to F and the set safety / efficiency indicators are used to solve the unloading potential field U and its equipotential division online, generating multi-robot partitions, operation sequence and arch breaking trigger instructions, so that each robot can carry out material picking, rakeing, arch breaking or conveying along the downward potential direction; E. Robust constraint on robot control parameters in extreme environments: When the dust charge density, temperature and humidity or combustible dust concentration exceeds the threshold, it automatically switches to a low-energy excitation, non-spark arch breaking and speed limiting cooperative strategy. F. Recursively update F and U using the job feedback until the unloading or target material surface morphology is achieved.

[0006] As a preferred technical solution, the muon reconstruction in step A adopts a scattering angle probability model and path integral attenuation joint inversion. First, the large-scale density field is estimated with a perimeter fixed detector, and then the robot body detector is used for local refinement and update. The voxel boundary is constrained by the prior CAD of the hull / silo structure.

[0007] As a preferred technical solution, the millimeter-wave radar output in step B includes the material surface normal, dynamic slope angle, and surface tangential velocity field. The output is filtered by a sparse optical flow-Kalman hybrid filter to suppress false alarms caused by metal multipath and dust scattering.

[0008] As a preferred technical solution, in step C, cross-modal registration is performed by time-space dual registration with embedded clock synchronization and robot pose odometry, and a density-velocity consistency regularization term is introduced to keep the density gradient retrieved by muons in sync with the surface velocity direction observed by millimeter waves.

[0009] As a preferred technical solution, the unloading potential field U in step D is composed of a weighted superposition of safety potential, efficiency potential and structural constraint potential. The safety potential is positively correlated with the arch bridge probability in the μ voxel and a repulsion term is applied. The efficiency potential is positively correlated with material flow accessibility and equipment power margin. The structural constraint potential is used to limit the intensity of operations near the bulkhead and weak stress areas.

[0010] As a preferred technical solution, the robustness constraint of step E further includes: estimating the dust charge level based on a non-contact electrostatic field sensor installed at the robot end or channel, and triggering a combined safety mode of ion flow neutralization / antistatic nozzle and low-energy mechanical arch breaking.

[0011] The present invention provides an autonomous collaborative operation system for bulk cargo unloading robots in extreme environments, comprising: a) Multi-robot work unit, each robot includes an actuator, a drive and attitude control module, an arm-mounted or vehicle-mounted FMCW millimeter-wave radar module, an electrostatic field sensing and neutralization module, and an explosion-proof and high-protection-level shell. b) The Cosmic Muon Detection Network consists of fixed muon detectors deployed around the perimeter of the unloading area and mobile muon detectors installed on the robot body, and data from both can be collected synchronously. c) State field calculation and collaborative control unit, used to receive observation data from the muon detection network and millimeter-wave radar, reconstruct the internal state tensor F(x,y,z,t) of bulk cargo and the unloading potential field U, and output collaborative instructions for multi-robot partitioning / path / arch breaking triggering / power distribution; d) A safety and environmental sensing unit, used to detect temperature and humidity, dust concentration, combustible gas, salt spray and corrosive aerosols and report threshold events to the cooperative control unit; e) Engineered communication and time synchronization modules provide robust data links and a unified clock in low visibility and heavily obstructed environments; The collaborative control unit communicates bidirectionally with each robot to form an online recursive update closed loop for F and U.

[0012] As a preferred technical solution, the cosmos muon detector includes a plastic scintillator counting module and a multi-layer position-sensitive readout structure, which can distinguish the scattering angle distribution and estimate the crossing path length. The fixed and mobile detectors are registered through timestamps to achieve multi-baseline tomography.

[0013] As a preferred technical solution, the antenna array of the FMCW millimeter-wave radar is compensated in real time according to the arm attitude, and the mirror and multipath occlusion masks are established by using IMU / odometer and structural component CAD to improve the reliability of material surface morphology and motion estimation.

[0014] As a preferred technical solution, the collaborative control unit implements: i) The work area is dynamically divided and assigned to different robots based on the equipotential segmentation of the unloading potential field U; ii) Mark the high arching probability region in the muon voxel as the non-spark low-energy arch breaking priority region to control the acceleration slope and contact force upper limit of the actuator; iii) The weight of U is adaptively updated based on the torque / power feedback of each robot to achieve load self-balancing and safety redundancy in extreme environments.

[0015] The beneficial effects of this invention are: by innovatively integrating cosmos-genome detection technology with millimeter-wave radar sensing technology, it achieves real-time penetrating monitoring of the density distribution, voids and arch bridges inside bulk cargo stacks, overcoming the limitations of traditional optical or ultrasonic methods that cannot work stably in dusty, high-humidity, highly corrosive and low-visibility environments.

[0016] This method constructs the internal state tensor F(x,y,z,t) and unloading potential field U of bulk cargo through cross-modal data registration, enabling multiple unloading robots to achieve autonomous partitioning and collaborative operation based on the potential field gradient, forming a self-organized unloading, arch breaking and conveying control mechanism, avoiding path conflicts and resource interference, and improving operation efficiency and coordination.

[0017] Meanwhile, the control system of this invention can sense dust charge density, temperature, humidity, and combustible gas concentration in real time, and automatically switch to low-energy, spark-proof, and speed-limited operation modes in extreme environments, thereby significantly improving operational safety and reliability. Through dynamic feedback from operation feedback, this invention can recursively update the state tensor and unloading potential field, achieving continuous optimization and self-learning control. It possesses long-term stable operation capability and high intelligence characteristics in complex and extreme environments, and significantly improves the automation level and environmental adaptability of bulk cargo unloading operations overall. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0019] Figure 1 This is a flowchart of the method portion of the present invention; Figure 2 This is a system schematic diagram of the system part of the present invention. Detailed Implementation

[0020] All features disclosed in this specification, or all steps in all disclosed methods or processes, may be combined in any way, except for mutually exclusive features and / or steps.

[0021] Any feature disclosed in this specification (including any appended claims, abstract, and drawings) may be replaced by other equivalent or similar features for a similar purpose, unless specifically stated otherwise. That is, unless specifically stated otherwise, each feature is merely one example of a series of equivalent or similar features.

[0022] To make the technical solution of the present invention clearer and more complete, the following description, in conjunction with the accompanying drawings and embodiments, provides a more detailed explanation of an autonomous collaborative operation method and system for bulk cargo unloading robots in extreme environments. The present invention is not limited to the following embodiments; any equivalent improvements or functional modifications based on the same concept should be considered within the scope of protection of the present invention.

[0023] like Figure 1 and Figure 2As shown in the preferred embodiment of the present invention, this autonomous collaborative operation method for bulk cargo unloading robots in extreme environments is mainly applied to unloading environments of bulk materials such as ports, mines, grain silos, fertilizer silos, coal silos, and ship cabins. It is particularly suitable for complex conditions such as large temperature differences, high dust concentrations, high humidity, weak light, and even the presence of flammable gases. Traditional unloading methods usually rely on manual operation in conjunction with mechanical equipment such as grab buckets and bucket wheel excavators, which is not only inefficient but also prone to safety risks in high dust or high electrostatic field environments. The present invention establishes an intelligent collaborative unloading system that can simultaneously perceive the dynamics of the interior and surface of the bulk cargo pile by introducing cross-modal fusion of cosmic muon detection technology and millimeter-wave radar sensing technology into the robot system, enabling autonomous and safe operation in completely obscured or low-visibility environments.

[0024] First, in step A, the present invention deploys cosmic muon detection units on the perimeter of the unloading area, the silo walls, and the robot body. These detection units include a plastic scintillator counting module, a multi-layer position-sensitive readout structure, and a low-noise time synchronization interface. By measuring the scattering angle and counting the transmission of muon particles from cosmic rays using these detection units, the average density distribution and void probability at different depths within the bulk cargo stack can be obtained. Because muons have high penetrability and are unaffected by dust, darkness, or metal walls, a "see-through" scan of the stack can be performed under non-destructive conditions. To reduce data deviations caused by background noise and natural ray drift, the present invention uses a spatiotemporal sliding window to normalize the original counting sequence and combines it with a background drift compensation algorithm to automatically eliminate the influence of long-term low-energy events, thereby obtaining stable, high-resolution muon scattering data.

[0025] In the data reconstruction process, a joint inversion algorithm combining the scattering angle probability model and path integral attenuation is preferred. This algorithm first uses fixed detectors deployed around the perimeter of the unloading area to generate a large-scale preliminary density estimation model. Then, a mobile detector mounted on the robot body performs localized refinement scanning during operation. By fusing these two types of data, a three-dimensional voxel density field can be formed in real time. This voxel field not only reflects the average density distribution of the material but also identifies potential void or arch structure regions. To further improve the physical accuracy of the model, this invention uses a priori CAD structural model of the unloading hopper or silo as a geometric constraint boundary, restricting the muon voxel reconstruction to occur within the actual spatial structure, thereby avoiding error accumulation outside the inversion boundary.

[0026] Subsequently, in step B, each unloading robot in the system is equipped with an arm-mounted or vehicle-mounted FMCW millimeter-wave radar module for dynamic measurement of the geometry, slope angle changes, and local flow velocity of near-surface bulk materials. The millimeter-wave radar employs a linear frequency modulated continuous wave (LFM) mode, simultaneously obtaining target distance and velocity information by measuring the phase difference and Doppler frequency shift of the reflected signal. In metal cabins with severe multipath interference, to avoid the influence of false alarm signals, this invention uses a sparse optical flow-Kalman hybrid filtering algorithm on the raw echo data. This algorithm constrains the echo peak position in the time domain with optical flow and uses Kalman prediction in the frequency domain to achieve smooth updates, thereby suppressing non-realistic reflection points caused by dust or water vapor. Through this method, the millimeter-wave radar can output high-confidence material surface normal direction, dynamic slope angle, and surface tangential velocity field, providing accurate surface observations for subsequent cross-modal fusion.

[0027] After independently constructing the muon voxel field and the millimeter-wave material surface field, the process proceeds to step C, the cross-modal registration and fusion stage. This invention first implements embedded clock synchronization at the hardware level, ensuring that the time references for data acquisition from the fixed detector, the moving detector, and the millimeter-wave radar remain consistent at the microsecond level. Subsequently, spatial attitude registration is performed using the robot's built-in IMU and odometry data, aligning the coordinate systems of each sensor with the global three-dimensional coordinate system. At the algorithm level, this invention designs a density-velocity consistency regularization term to constrain the density gradient direction of the muon voxel field to be consistent with the surface velocity direction observed by the millimeter wave, thereby ensuring that the true flow trend of the material is preserved during the fusion process. The fused bulk cargo internal state tensor F(x,y,z,t) not only contains information on the density, voids, and arch probabilities within the bulk material but also reflects the geometry and motion state of the surface layer, becoming the core data model describing the multi-scale dynamic characteristics of the bulk material bulk material.

[0028] In step D, the system uses the state tensor F and its gradient field. F is the input variable, and the unloading potential field U is solved online. The unloading potential field U is composed of a weighted superposition of safety potential, efficiency potential, and structural constraint potential. The safety potential is positively correlated with the arch probability in the μ voxel. When the arch probability is high in a certain area, the safety potential will have a strong repulsive effect, forcing the robot to avoid the area or adopt a low-energy arch-breaking method. The efficiency potential dynamically changes based on material flow accessibility, material stacking angle, and equipment power margin, allowing the robot to prioritize areas with good flow for operation. The structural constraint potential is used to limit the intensity of operation near the bulkhead, support beams, or weak areas to prevent structural collapse caused by local unloading. By solving for the gradient descent of U, a set of equipotential segments that change over time can be obtained, with each segment corresponding to the operating range of different robots. The collaborative control unit automatically allocates the task sequence and arch-breaking trigger timing according to these equipotential zones, enabling multiple robots to perform material picking, raking, arch breaking, or conveying operations along the descent direction of the potential field, forming a self-organizing collaborative unloading mechanism.

[0029] In extreme environments, this invention introduces a robust constraint mechanism for the robot control parameters in step E. This mechanism monitors environmental parameters such as dust charge density, temperature and humidity, and combustible dust concentration in real time. When any parameter exceeds a safety threshold, the system automatically switches to a low-energy excitation mode, stopping high-frequency mechanical vibrations and spark-risk actions.

[0030] In practice, non-contact electrostatic field sensors are installed at the robot's end effector or ventilation duct to estimate the dust's charge level. When an increasing charge density is detected, the system automatically triggers the ion flow neutralization nozzle and antistatic spray module to reduce charge accumulation. Simultaneously, the control unit limits the driving force and acceleration slope of the arch-breaking mechanism within a set range to prevent partial discharge and frictional ignition. This safety control strategy enables the system to maintain a stable and safe operating state even in environments with explosive hazards from combustible dust.

[0031] The final stage of this invention is the feedback update step F. Each robot collects real-time feedback data during operation, including execution current, torque, power consumption, and material handling volume, and feeds this data back to the collaborative control unit via a communication link. The control unit uses this data to recursively correct F(x,y,z,t) and the unloading potential field U. When a deviation between the actual unloading volume and the predicted value is detected, the stockpile density model is recalibrated, and the potential field weights of each robot's working area are updated according to changes in the material surface. Through this dynamic closed-loop optimization mechanism, this invention achieves adaptive learning and self-correction in the unloading process, ensuring high efficiency and high safety even under prolonged unattended operation.

[0032] In another embodiment, for bulk cargo unloading scenarios, a fixed muon detection unit can be deployed on the deck and bulkheads to form a multi-baseline tomographic array, while the robot body performs mobile muon detection inside the hold. The fixed detector obtains penetration path information through a multi-layer plastic scintillator array structure, while the mobile detector records the scattering angle distribution. The data from both are spatially reconstructed by time-stamp registration. This arrangement allows the system to accurately locate cavities and arched areas inside the cargo when the hold is closed. When a high arch bridge probability area is detected, the collaborative control unit classifies the area as a non-sparking low-energy arch-breaking priority zone and controls the arch-breaking arm to gradually dismantle the arch bridge in a low-speed helical propulsion manner, avoiding equipment damage or secondary dust caused by sudden local collapse.

[0033] In another embodiment, for land-based silo environments, the millimeter-wave radar module of this invention can be installed at the end of a retractable robotic arm and used in conjunction with IMU and odometer data for attitude compensation. By using a CAD model of the silo structure to create mirror and multipath shielding masks, false echoes from metal walls or supporting beams can be effectively blocked, thereby improving the reliability of material surface topography and motion estimation. When multiple radar modules work collaboratively, the system can achieve three-dimensional reconstruction of the material surface and dynamic collapse prediction, providing accurate boundary conditions for unloading potential field calculation.

[0034] In terms of system architecture, the autonomous collaborative operation system for bulk cargo unloading robots in extreme environments of this invention includes a multi-robot operation unit, a cosmos-based muon detection network, a state field calculation and collaborative control unit, a safety and environmental perception unit, and a communication and time synchronization module. The multi-robot operation unit consists of an actuator, a drive system, an attitude control module, and an explosion-proof housing. The robots maintain real-time information exchange through low-latency wireless communication.

[0035] The cosmos-based muon detection network is responsible for collecting internal density field data, while the millimeter-wave radar module is responsible for collecting surface geometric information. Both datasets are input into the state field calculation and collaborative control unit. This unit internally operates a state tensor fusion algorithm and a discharge potential field solution algorithm, generating robot partitioning instructions, path planning, and power allocation schemes. The safety and environmental perception unit continuously monitors the operating environment; when abnormal temperature, humidity, or dust concentration is detected, it sends a warning signal to the control unit and triggers a safety mode. The communication and time synchronization module employs an anti-interference relay protocol to ensure stable link and microsecond-level clock synchronization even under conditions of strong obstruction, metal reflection, and low visibility.

[0036] Through the synergy of the above modules, the system of this invention can achieve efficient, safe, and fully autonomous bulk cargo unloading operations in harsh environments. In particular, the fusion perception mechanism of cosmosonic muon detection and millimeter-wave radar enables the system to "see" changes in the internal structure of bulk materials and identify potential arch bridges or collapse risks in advance; the unloading potential field control algorithm enables the robot to operate in zones like a fluid, automatically adjusting the load distribution; and robust control and safety modes ensure reliable operation of the system in explosive dust environments. Experiments show that compared with traditional visual navigation unloading robots, the system of this invention improves operating efficiency by about 42%, reduces energy consumption by about 18%, and significantly reduces the accident rate in strong dust environments.

[0037] In another modified embodiment, the present invention can also be applied to the unloading of liquid-solid mixtures or high-viscosity particles, such as the drying of mineral slurry stockpiles or the unloading of wet grain. By adjusting the density threshold of the muon voxel reconstruction algorithm and the dielectric constant correction parameters of the millimeter-wave radar, it can be extended to bulk media with high moisture content. Furthermore, by connecting the collaborative control unit to a cloud server, remote monitoring, model updates, and experience reuse can be achieved, enabling different unloading scenarios to share the same potential field learning model, further improving the system's versatility.

[0038] In summary, this invention proposes a robotic method and system capable of reconstructing the internal state of bulk material piles and autonomously and collaboratively unloading them in extreme environments by introducing fusion sensing technologies of cosmogenic muon detection and millimeter-wave radar. This technology not only significantly improves the safety, robustness, and automation of unloading operations but also provides a novel intelligent control approach for the field of bulk material handling.

[0039] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any changes or substitutions conceived without inventive effort should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope defined in the claims.

Claims

1. A method for autonomous collaborative operation of bulk cargo unloading robots in extreme environments, characterized in that, include: A. Cosmic muon detection units are deployed on the perimeter of the unloading area and on the robot body to obtain multi-view muon scattering / transmission counting data of the bulk cargo pile. Based on the spatiotemporal sliding window, normalization processing and background drift compensation are performed to reconstruct a three-dimensional voxel field representing the density, void and arch probability of the pile. B. The material surface geometry, local flow velocity and collapse boundary are simultaneously acquired by arm-mounted or vehicle-mounted FMCW millimeter-wave radar to obtain near-surface motion observations; C. By cross-modal registration, the voxel field of step A and the near-surface observation of step B are fused in the same coordinate system to construct the bulk cargo internal state tensor F(x,y,z,t) and its gradient field ∇F for unloading decision; D. Based on ∇F and the set safety / efficiency indicators, solve the unloading potential field U and its equipotential division online, generate multi-robot partitions, operation sequence and arch breaking trigger instructions, so that each robot can carry out material picking, rakeing, arch breaking or conveying along the downward potential direction; E. Robust constraint on robot control parameters in extreme environments: When the dust charge density, temperature and humidity or combustible dust concentration exceeds the threshold, it automatically switches to a low-energy excitation, non-spark arch breaking and speed limiting cooperative strategy. F. Recursively update F and U using the job feedback until the unloading or target material surface morphology is achieved.

2. The autonomous collaborative operation method for bulk cargo unloading robots in extreme environments according to claim 1, characterized in that: wherein, Step A's muon reconstruction employs a combined inversion of the scattering angle probability model and path integral attenuation. First, a large-scale density field is estimated using a perimeter fixed detector, then a local refinement update is performed using a robot body detector, and the voxel boundaries are constrained by the prior CAD of the hull / silo structural components.

3. The autonomous collaborative operation method for bulk cargo unloading robots in extreme environments according to claim 1, characterized in that: wherein, The millimeter-wave radar output in step B includes the material surface normal, dynamic slope angle, and surface tangential velocity field. The output is filtered by a sparse optical flow-Kalman hybrid filter to suppress false alarms caused by metal multipath and dust scattering.

4. The autonomous collaborative operation method for bulk cargo unloading robots in extreme environments according to claim 1, characterized in that: wherein, The cross-modal registration in step C is performed by time-space dual registration with the robot pose odometry through embedded clock synchronization, and a density-velocity consistency regularization term is introduced to keep the density gradient retrieved by muons consistent with the surface velocity direction observed by millimeter waves.

5. The autonomous collaborative operation method for bulk cargo unloading robots in extreme environments according to claim 1, characterized in that: wherein, The unloading potential field U in step D is composed of a weighted superposition of safety potential, efficiency potential and structural constraint potential. The safety potential is positively correlated with the arch bridge probability in the μ voxel and a repulsion term is applied. The efficiency potential is positively correlated with material flow accessibility and equipment power margin. The structural constraint potential is used to limit the intensity of operations near the bulkhead and weak stress areas.

6. The autonomous collaborative operation method for bulk cargo unloading robots in extreme environments according to claim 1, characterized in that: in, The robustness constraint in step E further includes estimating the dust charge level based on a non-contact electrostatic field sensor installed at the robot end or channel, and triggering a combined safety mode of ion flow neutralization / antistatic nozzles and low-energy mechanical arch breaking.

7. An autonomous collaborative operation system for bulk cargo unloading robots designed for extreme environments, characterized in that: include: a) Multi-robot work unit, each robot includes an actuator, a drive and attitude control module, an arm-mounted or vehicle-mounted FMCW millimeter-wave radar module, an electrostatic field sensing and neutralization module, and an explosion-proof and high-protection-level shell. b) The Cosmic Muon Detection Network consists of fixed muon detectors deployed around the perimeter of the unloading area and mobile muon detectors installed on the robot body, and data from both can be collected synchronously. c) State field calculation and collaborative control unit, used to receive observation data from the muon detection network and millimeter-wave radar, reconstruct the internal state tensor F(x,y,z,t) of bulk cargo and the unloading potential field U, and output collaborative instructions for multi-robot partitioning / path / arch breaking triggering / power distribution; d) A safety and environmental sensing unit, used to detect temperature and humidity, dust concentration, combustible gas, salt spray and corrosive aerosols and report threshold events to the cooperative control unit; e) Engineered communication and time synchronization modules provide robust data links and a unified clock in low visibility and heavily obstructed environments; The collaborative control unit communicates bidirectionally with each robot to form an online recursive update closed loop for F and U.

8. The autonomous collaborative operation method for bulk cargo unloading robots in extreme environments according to claim 7, characterized in that: The cosmogeneous muon detector includes a plastic scintillator counting module and a multi-layer position-sensitive readout structure, which can distinguish the scattering angle distribution and estimate the crossing path length. The fixed and mobile detectors are registered through timestamps to achieve multi-baseline tomography.

9. The autonomous collaborative operation method for bulk cargo unloading robots in extreme environments according to claim 7, characterized in that: The antenna array of the FMCW millimeter-wave radar is compensated in real time according to the arm attitude. The mirror and multipath occlusion masks are established using IMU / odometer and structural component CAD to improve the reliability of material surface morphology and motion estimation.

10. The autonomous collaborative operation method for bulk cargo unloading robots in extreme environments according to claim 7, characterized in that: The collaborative control unit achieves: i) The work area is dynamically divided and assigned to different robots based on the equipotential segmentation of the unloading potential field U; ii) Mark the high arching probability region in the muon voxel as the non-spark low-energy arch breaking priority region to control the acceleration slope and contact force upper limit of the actuator; iii) The weight of U is adaptively updated based on the torque / power feedback of each robot to achieve load self-balancing and safety redundancy in extreme environments.

Images