Full-automatic dam slag salvaging pipe control system based on sensor fusion

Abstract

The application discloses a kind of based on sensor fusion's full-automatic dam slag salvaging pipe control system, belong to the technical field of intelligent operation and maintenance of hydropower station.System includes multi-source perception module, fusion decision module, integrated execution module, safety protection module and man-machine interaction module.Through water level, load, attitude, vision, wind speed, anti-collision multi-sensor fusion, realize environmental full-factor perception, automatically identify floating object type and adapt to water level and attitude balance, complete slag, log cutting, grappling hook salvage, automatic dumping, electric extension and manned safety operation, can realize eight-way movement with dam crane, one machine multiple hole operation.System has multistage safety protection, salvage load is greater than or equal to 1.5t, vertical stroke is greater than or equal to 9m, uses 380V industrial power supply, durable and reliable structure.The application realizes dam slag unmanned, intelligent, integrated operation, high efficiency, strong safety, low modification cost, suitable for various hydropower station floating object cleaning operation and maintenance.

Description

Technical Field

[0001] This invention relates to the field of intelligent operation and maintenance equipment technology for hydropower stations, specifically to a fully automated dam slag removal and control system based on sensor fusion. Background Technology

[0002] During the daily operation of hydropower station dams, garbage, floating debris, and driftwood in the river channel can easily accumulate in large quantities at the water intake, ramp entrance, and trash rack area. This can not only cause the trash rack to become clogged and the water intake of the units to be obstructed, but also cause safety hazards such as motor jamming and overload. At the same time, it can damage the river's ecological environment and increase the cost of manual operation and maintenance and operational risks.

[0003] The existing dam slag removal equipment has the following main technical defects: 1. The operation is mainly manual, relying on the operator's experience to judge the slag condition, water level and equipment attitude, resulting in low automation and high safety risks in high-altitude and water-related operations; 2. The sensing methods are limited, using only simple limit or load monitoring, which cannot achieve the integration of multi-dimensional information such as water level, attitude, wind speed, slag thickness and environmental images, making it difficult to adapt to complex working conditions; 3. The functions are limited, only having basic slag removal capabilities, unable to cut or grab large driftwood, and slag removal, personnel transport and extension operations are independent, resulting in low equipment utilization; 4. Poor compatibility with the existing dam gantry crane, unable to achieve multi-port mobile operation, resulting in high modification and supporting costs; 5. Lack of a complete heavy-load safety protection system, with high risks of overloading, overturning, falling and collision.

[0004] In summary, there is an urgent need for a fully automated dam slag removal management system based on sensor fusion, which can integrate multi-sensor fusion, fully automated closed-loop control, slag removal, cutting, grabbing, and manned operation, as well as dam crane collaborative operation, to meet the needs of hydropower stations for efficient, safe, intelligent, and low-cost slag removal operation and maintenance. Summary of the Invention

[0005] To address the aforementioned problems, this invention provides a fully automated dam slag removal and control system based on sensor fusion, which enables intelligent identification of floating objects, adaptive water level control, self-balancing of attitude, automatic switching of multiple functions, and full-process safety protection, thus solving the problems of low efficiency, high risk, limited functionality, and poor adaptability of traditional slag removal operations.

[0006] The technical solution of this invention is: a fully automated dam slag removal and control system based on sensor fusion, comprising a multi-source sensing module, a fusion decision-making module, an integrated execution module, a safety protection module, and a human-machine interaction module; the multi-source sensing module, the integrated execution module, the safety protection module, and the human-machine interaction module are all electrically connected to the fusion decision-making module; the multi-source sensing module is used to collect water level, load, attitude, distance, visual images, wind speed, and collision avoidance signals in real time; the fusion decision-making module is used to perform fusion processing on the multi-source sensing data to realize floating object classification, operation path planning, mode switching, and control command output; The integrated execution module is used to perform actions such as slag removal, slag cutting, hook retrieval, slag dumping, platform deformation, electric extension, personnel lifting, and gantry crane movement; the safety protection module is used to achieve graded protection against overload, overturning, collision, fall, and emergency shutdown; the human-machine interaction module is used for remote control, real-time image transmission, local monitoring, and emergency operation.

[0007] Furthermore, the multi-source sensing module includes a water level sensor, a load sensor, an attitude sensor, a vision sensor, a distance sensor, a wind speed sensor, and an anti-collision sensor. The water level sensor is used to monitor the vertical travel from the dam crest to the water surface, with an effective working travel of not less than 9m. The load sensor is used to monitor the salvage load, with a rated salvage load of not less than 1.5t and a safety redundancy of 2t. The attitude sensor is used to monitor the platform's tilt angle and displacement to achieve self-balancing control. The vision sensor is used to acquire information on the distribution, type, and size of floating objects. The wind speed sensor is used to collect environmental wind load data to provide compensation for attitude stability control.

[0008] Furthermore, the fusion decision module includes an edge computing unit, a data fusion algorithm unit, a pattern recognition unit, a path planning unit, and a control output unit; the data fusion algorithm unit fuses water level, load, attitude, visual, and wind speed data to establish a linkage model of scum thickness-salvage load-water depth; the pattern recognition unit is used to identify ordinary scum, large driftwood, debris attached to trash racks, and automatically match the operation mode.

[0009] Furthermore, the integrated execution module includes a dam crane coordinated electric moving mechanism, a platform self-deformation mechanism, an electric extension platform, an automatic slag dumping mechanism, a disc rotary cutting mechanism, a hanging hook mechanism, and a personnel safety platform. The dam crane coordinated electric moving mechanism has eight-way electrically controlled movement function, enabling multi-hole operation with one machine. The personnel safety platform can accommodate 1-2 people, is equipped with a 1.2m high safety fence and safety rope mounting position, and operates smoothly, with speed limit and anti-sway. The electric extension platform adopts electrified control and can expand the operating space.

[0010] Furthermore, the safety protection module includes a multi-level safety chain lock, a fall prevention mechanism, a wheel locking mechanism, an overload stop mechanism, an attitude over-limit protection mechanism, and an emergency stop unit; the multi-level safety chain lock is used to share the heavy load bearing capacity and prevent overload of the mounting nodes and structural parts from falling.

[0011] Furthermore, the system is powered by 380V industrial electricity with a connection distance of no less than 50m, meeting the waterproof and moisture-proof requirements of hydropower stations; the main structure is made of material of no less than Q235, which is corrosion-resistant and wear-resistant, with a design service life of no less than 10 years; the whole system adopts a modular and lightweight design, and can be hoisted and disassembled through its own structure.

[0012] Furthermore, the operating modes include automatic slag removal mode, driftwood cutting mode, hook retrieval mode, manned operation mode, and extended platform mode, supporting fully automatic switching and manual one-click switching.

[0013] A fully automated dam slag removal control method based on sensor fusion: S1. The system performs a power-on self-test, initializes the multi-source sensing module, and moves the dam crane's coordinated electric mobile mechanism to the designated working position. S2. The water level sensor collects real-time water level data, and the fusion decision module calculates the optimal lowering depth and initial attitude. S3: Visual and distance sensors collect information on floating objects, and the decision-making module identifies the type and distribution of floating objects. S4. The integrated decision-making module automatically matches the operation mode and drives the integrated execution module to complete the actions of retrieval, cutting, grabbing, dumping, and extending. S5. During operation, the safety protection module monitors the load, attitude, and collision avoidance signals in real time, and immediately triggers protection and shutdown in case of abnormal conditions. S6. When the operation is completed, the integrated execution module automatically resets, the gantry crane returns to its position, and the system is powered off and locked.

[0014] The beneficial effects of this invention are: 1. Fully automated unmanned operation: Through multi-sensor fusion, intelligent identification of slag conditions and full-process automatic control are achieved, greatly reducing reliance on manual labor and operational risks; 2. Strong adaptability to working conditions: Real-time sensing of water level, load, attitude, and wind speed, dynamically adjusting platform attitude and operation strategy to adapt to environments with large water level differences and strong wind loads. 3. Multifunctional integration: It integrates slag removal, cutting, grabbing, personnel carrying, extension, and automatic dumping functions, and one set of equipment meets the needs of multiple scenarios, improving utilization rate; 4. High safety: Multi-level safety chain lock, anti-fall, overload, overturning and anti-collision grade protection, stable and reliable operation under heavy load; 5. High adaptability and low cost: It can be directly connected to the existing gantry cranes of the dam, realizing eight-way movement and multi-span operation of one machine, without the need for new large-scale supporting facilities; 6. High versatility: Applicable to slag removal operations at various hydropower station dams, and has broad application value. Attached Figure Description

[0015] Figure 1 This is a schematic diagram of the multi-source sensing module structure of the present invention. Figure 2 This is a schematic diagram of the integrated execution module structure of the present invention. Figure 3 This is a schematic diagram of the workflow of the present invention. Detailed Implementation

[0016] The embodiments of the present invention will be further described below with reference to the accompanying drawings.

[0017] Example 1:

[0018] like Figure 1-3As shown, a fully automated dam slag removal and control system based on sensor fusion includes a multi-source sensing module, a fusion decision-making module, an integrated execution module, a safety protection module, and a human-machine interaction module. The multi-source sensing module includes a water level sensor, a load sensor, an attitude sensor, a vision sensor, a distance sensor, a wind speed sensor, and an anti-collision sensor. The water level sensor uses an immersion-type level transmitter with a measurement range of 0-15m, an accuracy of ±0.5%FS, and an output of a 4-20mA analog signal. It is installed on the sidewall of the dam's working orifice to collect real-time data from the dam crest. Vertical distance to the water surface, ensuring an effective working stroke ≥9m; Load sensor: a tension-type load cell with a rated load of 1.5t, a safe overload capacity of 2t, and an output of 0-5V voltage signal, integrated into the hoisting node of the slag removal mechanism, to monitor the slag removal load in real time, with trigger thresholds set at 1.8t (overload warning) and 2t (overload shutdown); Attitude sensor: a six-axis IMU inertial measurement unit, measuring range ±180° (tilt angle), ±500° / s (angular velocity), update frequency 100Hz, installed at the center of the slag removal platform, using... The system collects tilt, displacement, and angular velocity data from the acquisition platform to achieve self-balancing control within a ±5° range. A vision sensor, using a 2-megapixel industrial camera and LiDAR fusion module, operates at 30fps with a detection distance of 0.5-50m. Installed at the front of the platform, it uses the YOLOv8 algorithm to identify the type of floating debris (ordinary scum / large driftwood / debris attached to trash racks), size (≥0.5m is considered large driftwood), and distribution density. A distance sensor, using a laser rangefinder, measures from 0.1-100m with an accuracy of ±1mm. Installed at both ends of the overhead crane beam, it enables precise positioning of the crane's eight-way electrically controlled movement (positioning error ≤ 5cm); Wind speed sensor: a three-cup anemometer is selected, with a measurement range of 0-60m / s and an accuracy of ±0.2m / s. It is installed in the dam top working area. When the wind speed is ≥15m / s, the wind load compensation algorithm is triggered to reduce the platform's moving speed; Anti-collision sensor: an ultrasonic proximity switch is selected, with a detection distance of 0-2m and a response time ≤10ms. It is installed around the platform and at both ends of the overhead crane's moving track. After being triggered, it immediately decelerates (≤0.1m / s) or stops.

[0019] Specifically, the fusion decision-making module adopts an industrial-grade edge computing unit, equipped with a data fusion algorithm and a pattern recognition model. The edge computing unit uses an industrial-grade embedded controller (model: Intel Core i7-12700H), with 16GB of memory and 128GB of storage, supports 4G / 5G communication, and is pre-installed with a Linux operating system. It integrates data fusion algorithms, pattern recognition models, and path planning algorithms. Algorithm parameters: In the linkage model of scum thickness-salvage load-water depth, when the scum thickness is ≤0.3m, the water depth is 0.5m, and the upper limit of the salvage load is 1.5t; when the scum thickness is 0.3-1m, the water depth is 1m, and the upper limit of the salvage load is 1.2t; when the scum thickness is >1m, salvage is carried out in batches, with a single load upper limit of 1t.

[0020] Specifically, the integrated execution module includes an eight-way electric moving mechanism for the dam gantry crane, a platform self-deformation mechanism, an electric extension platform, an automatic tilting mechanism, a rotary disc saw, a hanging grab hook, and a personnel safety platform. The gantry crane electric moving mechanism is adapted to the existing 5t gantry crane at the dam, equipped with eight-way electrically controlled wheels, a moving speed of 0.5-2m / s (adjustable), a track length of 20m, covering 8 working positions, enabling "one machine, multiple positions" operation. The platform self-deformation mechanism adopts a hydraulically driven folding structure, with an unfolded platform size of 4m×2m and a folded size of 2m×1m, and an tilt angle adjustment range of 0-45°, adapting to operations with water level fluctuations of ±3m. Required: Electric extension platform: Utilizing an electric push-rod driven telescopic structure, with an extension length of 0-1.5m, a load capacity ≥500kg, and an extended working area covering 0.5-2m of the surrounding area of ​​the trash rack; Rotary disc cutting mechanism: Using a 500mm diameter alloy saw blade, a speed of 1500r / min, and a power of 3kW, capable of cutting driftwood with a diameter ≤0.8m; Personnel safety platform: Dimensions 2m×1.5m, equipped with a 1.2m high protective fence (3 horizontal bars), safety rope attachment points (4), and an anti-slip base, rated to carry 2 people (total weight ≤200kg), with a lifting speed ≤0.3m / s, and featuring anti-sway and anti-fall functions.

[0021] Specifically, the safety protection module includes a multi-level safety chain lock, fall arrest mechanism, wheel lock, overload shutdown, attitude over-limit protection, and emergency stop. The multi-level safety chain lock uses a double wire rope + mechanical locking structure with a breaking strength ≥10t and a redundancy coefficient ≥5. The fall arrest mechanism uses an electromagnetic braking fall arrestor with a braking distance ≤0.5m and a response time ≤0.1s. The emergency stop unit is equipped with a local emergency stop button (platform + gantry crane), a remote emergency stop controller (effective distance ≥100m), and a central control room emergency stop switch, providing three-level emergency stop linkage. The human-machine interface module uses a multi-channel industrial remote controller and a real-time video transmission unit. The system has a rated salvage load of 1.5t and a safety redundancy of 2t; a vertical working stroke of 9m; a 380V power supply with a connection distance of 50m; a main structure made of Q235 material with a design life of 10 years; and a manned platform accommodating two workers, equipped with a 1.2m safety fence.

[0022] Work process (taking Zangmu Hydropower Plant as an example): Step 1: System Initialization and Self-Test The operator starts the system in the central control room. The system automatically performs a power-on self-test: Multi-source sensing module: It checks the communication status of sensors such as water level, load, and attitude one by one. If there is an abnormality, it will alarm and display the fault location; Execution module: The gantry crane's electric moving mechanism is run unloaded for 50cm. The platform's self-deformation mechanism completes one folding-unfolding action to verify the mechanism's flexibility; Safety protection module: It triggers overload warning and anti-collision warning tests to verify the effectiveness of the protection mechanism; After the self-test is passed, the system enters the standby state, and the gantry crane automatically moves to the preset initial working hole position (such as hole #3).

[0023] Step 2: Environmental Perception and Operating Condition Assessment The water level sensor recorded the current water level as 7.8m, and the fusion decision module's computing platform was lowered to a depth of 8m (with a 0.2m safety distance reserved). The visual sensor and lidar collected images of the work area and identified approximately 20㎡ of mixed floating debris at borehole #3, including three pieces of driftwood with a diameter of 0.5-0.7m and a scum thickness of approximately 0.6m. The wind speed sensor recorded the current wind speed as 8m / s, requiring no wind load compensation. Based on the above data, the fusion decision module generated the following work strategy: first, salvage the ordinary scum, then cut the driftwood, and finally grab the cut driftwood.

[0024] Step 3: Fully Automated Slag Removal Operation Ordinary scum removal: The integrated decision-making module issues an instruction, and the gantry crane lowers the platform to 0.2m above the water surface. The platform's self-deformation mechanism adjusts the tilt angle by 15°, so that the scum bucket is submerged 0.8m in the water. The gantry crane's electric moving mechanism drives the platform to move laterally at a speed of 0.8m / s. The scum bucket collects the scum, and the load sensor monitors the load in real time. When the load reaches 1.2t, the collection stops. The platform is raised to the dam top height (9m), and the gantry crane moves above the scum truck. The scum dumping mechanism automatically flips the scum bucket to dump the scum into the scum truck. After dumping, it resets. Large driftwood processing: Visual sensors locate the driftwood position, the overhead crane moves the platform above the driftwood, and the electric extension platform extends 1m to get close to the driftwood; the disc rotating cutting mechanism is activated, the platform is slowly lowered, and the saw blade cuts into the driftwood in 3 cuts (each cut to a depth of 0.2m), cutting the driftwood into segments ≤0.3m; the hook mechanism is activated to grab the cut driftwood segments, lift them to the top of the dam, and then dump them into the designated area.

[0025] Step 4: Manned Operation Mode When manual inspection of the trash rack blockage is required, the operator switches to "manned operation mode" in the central control room: the system automatically locks the slag removal / cutting mechanism to prevent accidental operation; the personnel safety platform fence automatically locks, and the anti-slip base inflatable pad is inflated (to increase cushioning); two maintenance personnel enter the platform, are secured with safety ropes, and the operator uses a remote control to lower the platform to the trash rack position at a speed of 0.2 m / s; the electric extension platform extends 1.2 m, and after the maintenance personnel complete the cleaning and inspection of the trash rack, the platform is raised to the top of the dam at a speed of 0.3 m / s, the fence is unlocked, and the personnel evacuate.

[0026] Step 5: Safety Monitoring and Anomaly Handling Throughout the operation, the safety protection module monitors in real time: if the load sensor detects a load of 1.8t, the system issues an audible and visual alarm, indicating an "overload warning," and stops collecting scum; if the load reaches 2t, the system immediately stops and locks the platform, awaiting manual intervention; if the attitude sensor detects a platform tilt angle > 5°, the system automatically activates the self-balancing algorithm and adjusts the hydraulic mechanism to reset; if the tilt angle > 8°, the attitude over-limit protection is triggered, the system stops and locks; if the anti-collision sensor detects a distance of < 0.5m between the platform and the trash rack, the system immediately decelerates to 0.1m / s, and stops when the distance is < 0.2m.

[0027] Step 6: Job Completion and System Reset Once the work at hole #3 is completed, the vision sensor confirms that there are no residual floating objects, and the system automatically performs a reset: the platform's self-deformation mechanism folds, the electric extension platform retracts, and the cutting / grappling hook mechanism resets; the gantry crane moves to hole #4, and steps 2-5 are repeated to complete the work at all holes; after all work is completed, the gantry crane moves to the preset parking position, the platform is powered off, the safety locking mechanism locks, and the system enters standby mode.

[0028] Performance test results: The system underwent a 30-day field test at the Zangmu Hydropower Plant. The test results are as follows: Operational efficiency: It can complete the slag removal operation of 8 working holes in a single day, with a total slag removal volume of ≥15t, which is 3 times more efficient than traditional manual slag removal (≤5t per day); Safety: No safety accidents such as overloading, collisions, or falls occurred during the test. The time for manual high-altitude water wading operations was reduced from 8 hours per day to 0.5 hours (only for manned inspections), and the safety risk was reduced by 90%. Adaptability: Under conditions of water level fluctuation of ±2.5m and wind speed ≤12m / s, the platform's self-balancing control accuracy is ≤±1°, and its operational stability meets the standards. Reliability: The system operates for an average of 10 hours per day, with a failure rate of ≤0.5 times / day and a fault recovery time of ≤10 minutes.

[0029] The embodiments described above are merely illustrative of specific implementations of the present invention, and while the descriptions are detailed, they should not be construed as limiting the scope of the present invention. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these modifications and improvements all fall within the scope of protection of the present invention.

Claims

1. A fully automated dam slag removal and control system based on sensor fusion, characterized in that, include: Multi-source sensing module, fusion decision-making module, integrated execution module, security protection module, and human-computer interaction module; The multi-source sensing module, integrated execution module, security protection module, and human-computer interaction module are all electrically connected to the fusion decision module; The multi-source sensing module is used to collect water level, load, attitude, distance, visual images, wind speed, and collision avoidance signals in real time. The fusion decision module is used to fuse multi-source sensing data to achieve floating object classification, operation path planning, mode switching and control command output; The integrated execution module is used to perform actions such as slag removal, slag cutting, hook retrieval, slag dumping, platform deformation, electric extension, personnel lifting, and overhead crane movement. The safety protection module is used to achieve graded protection against overload, overturning, collision, fall, and emergency shutdown. The human-computer interaction module is used for remote control, real-time image transmission, local monitoring and emergency operation.

2. The system according to claim 1, characterized in that, The multi-source sensing module includes a water level sensor, a load sensor, an attitude sensor, a vision sensor, a distance sensor, a wind speed sensor, and an anti-collision sensor; the effective working distance of the water level sensor is not less than 9m; the rated salvage load of the load sensor is not less than 1.5t, and the safety redundancy is designed at 2t.

3. The system according to claim 1, characterized in that, The fusion decision module includes an edge computing unit, a data fusion algorithm unit, a pattern recognition unit, a path planning unit, and a control output unit; the data fusion algorithm unit establishes a linkage model of scum thickness, salvage load, and water depth.

4. The system according to claim 1, characterized in that, The integrated execution module includes a dam crane coordinated electric moving mechanism, a platform self-deformation mechanism, an electric extension platform, an automatic slag dumping mechanism, a disc rotation cutting mechanism, a hanging hook mechanism, and a personnel safety platform; the dam crane coordinated electric moving mechanism has eight-way electrically controlled movement function, realizing multi-hole operation with one machine.

5. The system according to claim 4, characterized in that, The manned safety platform can accommodate 1 to 2 people, and is equipped with a 1.2m high safety fence and safety rope attachment point. It operates smoothly, with speed limit and anti-sway.

6. The system according to claim 1, characterized in that, The safety protection module includes a multi-level safety chain lock, a fall protection mechanism, a wheel locking mechanism, an overload shutdown mechanism, an attitude over-limit protection mechanism, and an emergency stop unit.

7. The system according to claim 1, characterized in that, The system is powered by 380V industrial electricity, with a connection distance of not less than 50m, meeting the waterproof and moisture-proof standards of hydropower stations; the main structure is made of materials of not less than Q235.

8. The system according to claim 1, characterized in that, The operating modes include automatic slag removal mode, driftwood cutting mode, hook retrieval mode, manned operation mode, and extended platform mode, supporting fully automatic switching and manual one-click switching.

9. A fully automated dam slag removal and control method based on sensor fusion, characterized in that, Includes the following steps: S1. The system performs a power-on self-test, initializes the multi-source sensing module, and moves the dam crane's coordinated electric mobile mechanism to the designated working position. S2. The water level sensor collects real-time water level data, and the fusion decision module calculates the optimal lowering depth and initial attitude. S3: Visual and distance sensors collect information on floating objects, and the decision-making module identifies the type and distribution of floating objects. S4. The integrated decision-making module automatically matches the operation mode and drives the integrated execution module to complete the actions of retrieval, cutting, grabbing, dumping, and extending. S5. During operation, the safety protection module monitors the load, attitude, and collision avoidance signals in real time, and immediately triggers protection and shutdown in case of abnormal conditions. S6. When the operation is completed, the integrated execution module automatically resets, the gantry crane returns to its position, and the system is powered off and locked.

Images