Robot navigation system and method for agaricus bisporus planting shelf

Abstract

The invention discloses a robot navigation system for a agaricus bisporus planting shelf, the agaricus bisporus planting shelf comprises a support column and a planting groove, and the robot navigation system comprises a robot body, a navigation module and a composite shelf terrain; the robot body comprises a robot body, a four-wheel driving module and a controller; the navigation module comprises a laser range finder array and an acceleration sensor; the laser range finder array comprises a plurality of laser range finder sensor modules installed on the robot body. The method solves the problems of GPS failure, visual occlusion and high deployment cost in a shelf planting scene, has the characteristics of high environmental adaptability, high robustness and accurate operation, and is suitable for automatic management of multilayer planting of the agaricus bisporus.

Description

Technical Field

[0001] The present invention relates to the field of agricultural robot navigation technology, and in particular to a robot navigation system and method for Agaricus bisporus cultivation shelves. Background Art

[0002] With the rapid development of factory farming, the three-dimensional rack cultivation model for edible fungi such as Agaricus bisporus has become mainstream. This type of cultivation scenario typically uses a multi-layer stacked rack structure, with standardized planting troughs, vertical support columns, and narrow passages (usually 0.6-1.0 meters wide) forming a gridded working space. However, traditional agricultural robots face the following technical bottlenecks in this scenario:

[0003] (1) GPS failure: The metal frame of the shelves and the closed factory structure seriously shield the GPS signal, making satellite positioning unavailable;

[0004] (2) Visual interference: The dense arrangement of planting troughs, uneven illumination between layers (illumination fluctuations of 500-2000 Lux), and texture similarity caused by hyphae coverage resulted in a visual SLAM (Simultaneous Localization and Mapping) matching error rate exceeding 40%;

[0005] (3) The cost of deploying magnetic tracks is high: the cost of laying magnetic tracks per square meter is about 200-300 yuan, and it is difficult to adapt to dynamic adjustments of shelves (such as changes in floor height and channel reorganization).

[0006] (4) Existing laser navigation solutions rely on reflectors or preset markers. However, in the shelf scenario, the regular arrangement of planting troughs and support columns is prone to multiple reflection interference (false detection rate > 15%), and the marker maintenance cost accounts for more than 20% of the total equipment cost.

[0007] (5) Ultrasonic and infrared sensors are easily affected by temperature and humidity in narrow channels (humidity in Agaricus bisporus cultivation environment is >85%), and the ranging error fluctuates up to ±10%. Summary of the Invention

[0008] Purpose of the invention: The present invention aims to provide a robot navigation system and method for Agaricus bisporus cultivation shelves. Through multi-sensor fusion and feature matching algorithms, the system solves the problems of GPS signal shielding, visual occlusion interference, and high deployment costs existing in traditional navigation technologies in shelf cultivation scenarios, thereby realizing autonomous path planning without external markers, precise intersection recognition, and parking control.

[0009] Technical solution: A robot navigation system for a Agaricus bisporus cultivation rack, wherein the Agaricus bisporus cultivation rack includes support columns and planting troughs, and the robot navigation system includes a robot body and a navigation module;

[0010] The robot body includes a robot body, a four-wheel drive module, and a controller;

[0011] The navigation module includes a laser rangefinder array and an acceleration sensor;

[0012] The laser rangefinder array includes a plurality of laser rangefinder sensors mounted on the robot body, wherein:

[0013] The forward navigation sensor is installed in front of the vehicle, pointing in the direction of travel, to detect changes in the length of the road ahead and obstacles;

[0014] The lateral deviation correction group sensors are symmetrically arranged on both sides of the vehicle body to measure the real-time distance between the sensor and the side wall of the planting trough;

[0015] The support column detection group sensor is installed in the center of both sides of the vehicle body. It triggers the parking position counting by capturing the pulse signal reflected by the support column;

[0016] The rear monitoring group sensor is installed at the rear of the vehicle to perform reversing path verification and retreat control.

[0017] Furthermore, the forward navigation group sensor is a first sensor installed in front of the vehicle, pointing in the direction of travel, and the first sensor is installed at a height of H1. <H base , H base is the height of the robot's top surface from the ground;

[0018] The rear monitoring group sensor is the second sensor installed directly behind the rear of the vehicle, pointing in the reverse direction, and the second sensor is installed at a height of H2 <H leg , H leg is the height of the support column, that is, the vertical distance from the ground to the planting trough;

[0019] The lateral deviation correction group sensors are the third to sixth sensors. The third and fourth sensors are respectively arranged on both sides of the vehicle body near the front position, and the fifth and sixth sensors are respectively arranged on both sides of the vehicle body near the rear position; the third to sixth sensors are all pointed in the direction perpendicular to the direction of travel, and the installation height H3 meets the following requirements: H shelfb ≤H3≤H shelfc , H shelfb H is the vertical height of the bottom of the planting trough from the ground. shelfc It is the vertical height of the top surface of the planting trough from the ground;

[0020] The support column detection group sensors are the seventh sensor and the eighth sensor, which are respectively arranged at the center of both sides of the vehicle body. The seventh sensor and the eighth sensor are installed at a height of H4. <H leg .

[0021] Furthermore, the robot navigation system is applied to composite shelf terrain, which is a terrain formed by stacking and arranging multiple Agaricus bisporus cultivation shelves on the ground. The roads under this terrain are always straight and the intersections are right-angle intersections.

[0022] A method for utilizing a robot navigation system for Agaricus bisporus cultivation shelves comprises the following steps:

[0023] Step 1: The robot is powered on and the controller initializes the system.

[0024] Step 2: The controller turns on the laser rangefinder array and accelerometer to collect environmental data;

[0025] Step 3: As the robot moves forward, the third to sixth sensors emit lasers to measure the distance to the planting slots of the Agaricus cultivation racks, calculate the robot's posture, and make real-time corrections to ensure straight movement.

[0026] Step 4: The controller runs the intersection detection algorithm and the stop detection algorithm in real time;

[0027] Step 5: After reaching the manually set end point, if the robot needs to return, the original backward direction is changed to the new forward direction, the recognition position of each sensor is reset, the forward direction of the robot is reversed, and then the process returns to step 3 until the user actively stops the robot or the number of round trips reaches the set value.

[0028] Furthermore, step 4 specifically includes:

[0029] Step 4.1: The controller runs the parking detection algorithm. If the parking condition is met, the robot will pause and wait for the parking time to end before resuming operation.

[0030] Step 4.2: The controller runs the intersection detection algorithm, detects the road type ahead, and selects a road to move forward according to the preset road selection strategy.

[0031] Furthermore, the road types include straight roads, crossroads, left L-shaped intersections, right L-shaped intersections, T-shaped intersections, left-turn intersections, and right-turn intersections.

[0032] Furthermore, the road selection strategy includes setting priorities for right turns, left turns, and going straight. (For example, the priorities for road selection from high to low are right turns, going straight, and left turns. Marked road branches and road branches with no access are not included in road selection.)

[0033] Furthermore, the intersection detection algorithm includes:

[0034] (1) Straight line judgment: When the measured value of the first sensor exceeds 1.5W corridor , Wcorridor The standard width of the corridor, the fluctuation range of the measurement values ​​of the third sensor and the fourth sensor is ≤0.1W corridor , it is determined to enter the straight road;

[0035] (2) Intersection judgment: When the measurement value of the first sensor continues to decrease, and the measurement values ​​of the third and fourth sensors are both greater than 1.5W corridor The controller determines that the vehicle has entered an intersection and pushes the intersection record onto the stack;

[0036] (3) Left L-shaped intersection judgment: When the first sensor exceeds 1.5W corridor , the third sensor measurement value suddenly increases to ≥2W corridor , while the fourth sensor measurement value fluctuation range ≤ 0.1W corridor , the controller determines that it has entered a left L-shaped intersection and pushes the intersection record into the stack;

[0037] (4) Right L-shaped intersection judgment: When the first sensor exceeds 1.5W corridor , the fourth sensor measurement value suddenly increases to ≥2W corridor , while the third sensor measurement value fluctuation range is ≤0.1W corridor , the controller determines it as a right L-shaped intersection and pushes the intersection record into the stack;

[0038] (5) T-junction judgment: When the measurement value of the first sensor continues to decrease, and the measurement values ​​of the third and fourth sensors are both greater than 1.5W corridor , the controller determines it as a T-junction and pushes the record of this intersection into the stack;

[0039] (6) Left turn intersection judgment: When the measurement value of the first sensor continues to decrease and the measurement value of the third sensor suddenly increases to ≥2W corridor , while the fourth sensor measurement value fluctuation range ≤ 0.1W corridor , the controller determines it as a left turn intersection and pushes the intersection record into the stack;

[0040] (7) Right turn intersection judgment: When the measurement value of the first sensor continues to decrease and the measurement value of the fourth sensor suddenly increases to ≥2W corridor , while the third sensor measurement value fluctuation range is ≤0.1W corridor , the controller determines it as a right turn intersection and pushes the intersection record into the stack;

[0041] (8) Blockage and fallback judgment: When the first sensor measurement value is ≤2W corridor , and the fluctuation range of the measurement values ​​of the third sensor and the fourth sensor is ≤0.1W corridor , stop and wait for 10 seconds. During the waiting period, if the measured value of the first sensor exceeds 1.5W corridorIf so, the robot continues to move forward. If the waiting time ends and the measurement value of the first sensor remains unchanged, the controller determines that the current road branch has been completed, marks this branch, retreats to the previous intersection, selects other road branches to move forward. If there are no other branches to choose from at the current intersection, the controller determines that all traversals of this intersection have been completed, records this intersection out of the stack, returns to the previous intersection of this intersection and continues to judge until the stack is empty or the end point is reached.

[0042] Further, the stopping algorithm includes:

[0043] (1) The controller collects the pulse signals of the layer rack laser sweeping across the support column through the seventh sensor and the eighth sensor. When the controller detects that the distance value suddenly changes to D leg ≤W corridor and the duration Δt is within the set range, it is defined as passing through a support column; D leg is the laser distance from the sensor emitting laser to the support column and returning, and W corridor is the standard width of the corridor;

[0044] (2) Establish a support column pulse counter, and the count value is C leg , and the initial value is the number of remaining support columns N target before the target stopping position;

[0045] (3) Every time an effective support column is passed, C leg =C leg -1. When C leg =0, the deceleration and stopping program is triggered, and the control law v(t) is:

[0046] v(t)=v0·e -λt

[0047] v0 is the maximum speed, λ is the deceleration coefficient, and t is the time;

[0048] (4) After stopping, calibrate the lateral distance from the lateral deviation correction group sensor to the planting trough to satisfy ∣D side -D set ∣<u, lock the wheels. Dside represents the distance between the lateral deviation correction group sensor and the planting trough, Dset is the set threshold, and u is the set target difference.

[0049] Further, in step 3, the third sensor to the sixth sensor measure the real-time distances from themselves to the side wall of the planting trough, and generate deviation correction instructions through the PID algorithm.

[0050] Beneficial effects: Environmental adaptability: Navigation is achieved entirely by relying on the inherent structure of the rack (planting troughs, support columns, and right-angle roads), without the need to deploy external markers, reducing costs by more than 60%; High robustness: Multi-sensor redundant design (such as forward + lateral collaborative verification of intersections) resists the risk of single sensor failure; Dynamic response: Intersection detection response time is <0.3s, and the support column pulse counting error rate is <0.5%. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] Figure 1 This is a schematic structural diagram of a robot according to an embodiment of the present invention;

[0052] Figure 2 This is a schematic diagram of a shelf and stacking structure according to an embodiment of the present invention;

[0053] Figure 3 This is a schematic diagram of the robot working according to an embodiment of the present invention;

[0054] Figure 4 This is a machine working flow diagram according to an embodiment of the present invention;

[0055] Figure 5 This is a scene topographic map according to an embodiment of the present invention. DETAILED DESCRIPTION

[0056] The technical solution of the present invention will be further described below with reference to the accompanying drawings.

[0057] like Figure 1 As shown, the present invention provides a robot navigation system for Agaricus bisporus cultivation shelves, the robot navigation system includes a robot body and a navigation module;

[0058] The robot body includes a robot body, a four-wheel drive module, and a controller;

[0059] The navigation module includes a laser rangefinder array and an acceleration sensor;

[0060] The laser rangefinder array includes a plurality of laser rangefinder sensors mounted on the robot body, wherein:

[0061] The forward navigation sensor is installed in front of the vehicle, pointing in the direction of travel, to detect changes in the length of the road ahead and obstacles;

[0062] The lateral deviation correction group sensors are symmetrically arranged on both sides of the vehicle body to measure the real-time distance between the sensor and the side wall of the planting trough;

[0063] The support column detection group sensor is installed in the center of both sides of the vehicle body. It triggers the parking position counting by capturing the pulse signal reflected by the support column;

[0064] The rear monitoring group sensor is installed at the rear of the vehicle to perform reversing path verification and retreat control.

[0065] Furthermore, the forward navigation group sensor is a first sensor 101 installed in front of the vehicle, pointing in the direction of travel, and the first sensor is installed at a height H1. <H base , H base is the height of the robot's top surface from the ground;

[0066] The rear monitoring group sensor is the second sensor 102 installed directly behind the rear of the vehicle, pointing in the reverse direction, and the second sensor is installed at a height of H2 <H leg , H leg is the height of the support column, that is, the vertical distance from the ground to the planting trough;

[0067] The lateral deviation correction group sensors are the third to sixth sensors, the third sensor 103 and the fourth sensor 104 are respectively arranged on both sides of the vehicle body near the front position, and the fifth sensor 105 and the sixth sensor 106 are respectively arranged on both sides of the vehicle body near the rear position;

[0068] The third to sixth sensors all point to the vertical direction of travel, and the installation height H3 meets the following requirements: H shelfb ≤H3≤H shelfc , H shelfb H is the vertical height of the bottom of the planting trough from the ground. shelfc It is the vertical height of the top surface of the planting trough from the ground;

[0069] The support column detection group sensors are the seventh sensor 107 and the eighth sensor 108, which are respectively arranged at the center of both sides of the vehicle body. The seventh sensor 107 and the eighth sensor 108 are installed at a height of H4. <H leg .

[0070] like Figure 2 As shown, the robot navigation system is applied to a composite shelf terrain, which is a terrain formed by stacking and arranging a plurality of Agaricus bisporus cultivation shelves on the ground. The roads under this terrain are always straight and the intersections are right-angle intersections.

[0071] A method for utilizing a robot navigation system for Agaricus bisporus cultivation shelves comprises the following steps:

[0072] Step 1: The robot is powered on and the controller initializes the system.

[0073] Step 2: The controller turns on the laser rangefinder array and accelerometer to collect environmental data;

[0074] Step 3: As the robot moves forward, the third to sixth sensors emit lasers to measure the distance to the planting slots of the Agaricus cultivation racks, calculate the robot's posture, and make real-time corrections to ensure straight movement.

[0075] Step 4: The controller runs the intersection detection algorithm and the stop detection algorithm in real time;

[0076] Step 5: After reaching the manually set end point, if the robot needs to return, the original backward direction is changed to the new forward direction, the recognition position of each sensor is reset, the forward direction of the robot is reversed, and then the process returns to step 3 until the user actively stops the robot or the number of round trips reaches the set value.

[0077] Furthermore, step 4 specifically includes:

[0078] Step 4.1: The controller runs the parking detection algorithm. If the parking condition is met, the robot will pause and wait for the parking time to end before resuming operation.

[0079] Step 4.2: The controller runs the intersection detection algorithm, detects the road type ahead, and selects a road to move forward according to the preset road selection strategy.

[0080] Figure 3 This is a schematic diagram of the robot's work; Figure 4 This is a machine working flow diagram according to an embodiment of the present invention;

[0081] Implementation of the navigation system of the present invention:

[0082] Implementation environment configuration:

[0083] Shelf parameters:

[0084] Support column height H leg =1.8m, height of planting trough bottom H shelfb =0.5m, height of planting trough top surface H shelfc =0.8m; standard corridor width W corridor =0.8m, and the shelves are arranged in a 5×5 grid of right-angle roads.

[0085] Robot parameters:

[0086] Chassis size 0.6m×0.4m, chassis top height H base =0.4m; four-wheel independent drive, maximum speed v0 = 0.5m / s, deceleration coefficient λ = 0.8.

[0087] Sensor installation:

[0088]

[0089] Navigation process instance:

[0090] Scenario description: The robot starts from starting point A (shelf grid coordinate 1-1), moves along a straight road to the target stop position B (between coordinates 3-4 and 3-5), and finally reaches the end point C (coordinate 3-1). It passes through the right L-shaped intersection and T-shaped intersection. The intersection selection strategy is set to select by priority. The priority is set from high to low as left turn > forward > right turn. The stop position C target Set to 3. The scene terrain map is as follows Figure 5 shown.

[0091] Process 1: After the robot is started and initialized, the lateral sensors 103-106 continuously measure the distance D between the left and right planting grooves. left =0.42m, D right = 0.38m, calculate the lateral deviation ΔD = |0.42-0.38| / 2 = 0.02m, which does not exceed the threshold value of 0.05m. During the operation of the robot, the sensors 107-108 always detect the support column pulse signal. Every time the robot passes a support column C leg +1;

[0092] Process 2: When driving straight, the acceleration sensor detects that the vehicle body is slightly tilted to the right (angle θ = 1.2°). The controller adjusts the speed of the right wheel to reduce by 0.1m / s, and balance is restored after 5 seconds.

[0093] Process 3: When the robot passes through coordinate 1-2 and moves to 1-3, the measurement value of the lateral sensor 103 suddenly increases to D side =4m>1.5W corridor (Actual 1.5×0.8=1.2m), it is determined to be a right L-shaped intersection, the intersection coordinates (1-2) are pushed into the navigation stack, and straight ahead is selected according to the preset priority;

[0094] Process 4: When reaching coordinates 1-5, sensor 101 detects the rate of change of the forward distance (lasts 3 seconds) and the lateral sensor fluctuation is ≤0.08m. The robot stops and waits for 10 seconds. During this period, the measurement value of sensor 101 does not change, and there are still nodes in the navigation stack. It is determined that the road branch has been traversed and it is necessary to return to the previous intersection. At this time, the robot starts to back away, returns to coordinates 1-2, chooses to go right, and pops coordinates 1-2 from the stack.

[0095] Process 5: When the robot passes through coordinates 2-3 and moves to 3-3, sensor 101 detects the rate of change of the forward distance. (lasts 3 seconds), the measurement value of the lateral sensor 103-104 suddenly increases to D side =4m>1.5W corridor (Actual 1.5×0.8=1.2m), it is determined to be a T-junction, the intersection coordinates (2-3) are pushed into the navigation stack, and the left lane is selected according to the preset priority;

[0096] Process 6: When the robot passes between 3-4 and 3-5, C leg =C target =3, triggering parking, initial speed v0 = 0.5m / s, deceleration according to v(t) = 0.5·e-0.8t, and the speed drops to 0.05m / s after 3 seconds; after completely stopping, the wheels are locked, and after the parking time is over, the robot continues to move forward.

[0097] Process 7: When reaching coordinate 3-5, sensor 101 detects the rate of change of the forward distance (lasts 3 seconds) and the lateral sensor fluctuation is ≤0.08m. The robot stops and waits for 10 seconds. During this period, the measurement value of sensor 101 does not change, and there are still nodes in the navigation stack. It is determined that the road branch has been traversed and it is necessary to return to the previous intersection. At this time, the robot starts to return to coordinates 2-3, chooses to go right, and pops coordinates 2-3 from the stack;

[0098] Process 8: After the robot reaches 3-1, it detects the end position and ends the mission.

Claims

1. A robot navigation system for a Agaricus bisporus cultivation rack, wherein the Agaricus bisporus cultivation rack comprises a support column and a cultivation trough, characterized in that The robot navigation system includes a robot body and a navigation module; The robot body includes a robot body, a four-wheel drive module, and a controller; The navigation module includes a laser rangefinder array and an acceleration sensor; The laser rangefinder array includes multiple laser rangefinder sensors installed on the robot body, among which: the forward navigation group sensor is installed in front of the front of the vehicle, pointing in the direction of travel, detecting changes in the length of the road ahead and obstacles; the lateral correction group sensor is symmetrically arranged on both sides of the vehicle body, measuring the real-time distance between the sensor and the side wall of the planting trough; the support column detection group sensor is installed in the center position on both sides of the vehicle body, triggering the parking position count by capturing the pulse signal reflected by the support column; the rear monitoring group sensor is installed at the rear of the vehicle to perform reversing path verification and retreat control.

2. A robot navigation system for Agaricus bisporus cultivation shelves according to claim 1, characterized in that: The forward navigation group sensor is a first sensor (101) installed in front of the vehicle head, pointing to the direction of travel, and the first sensor is installed at a height H1 <H base , H base The height of the robot body top surface from the ground; the rear monitoring group sensor is the second sensor (102) installed directly behind the rear of the vehicle, pointing to the reverse direction, the second sensor installation height H2 <H leg , H leg is the height of the support column, that is, the vertical distance from the ground to the planting trough; the lateral correction group sensors are the third to sixth sensors, the third sensor (103) and the fourth sensor (104) are respectively arranged on both sides of the vehicle body near the front position, the fifth sensor (105) and the sixth sensor (106) are respectively arranged on both sides of the vehicle body near the rear position; the third to sixth sensors are all pointed in the vertical direction of the travel direction, and the installation height H3 meets: H shelfb ≤H3≤H shelfc , H shelfb H is the vertical height of the bottom of the planting trough from the ground. shelfc is the vertical height of the top surface of the planting trough from the ground; the support column detection group sensors are the seventh sensor (107) and the eighth sensor (108), which are respectively arranged at the center position on both sides of the vehicle body, and the seventh sensor (107) and the eighth sensor (108) are installed at a height of H4 <H leg .

3. The robot navigation system for Agaricus bisporus cultivation shelves according to claim 1, characterized in that: The robot navigation system is applied to a composite shelf terrain, which is a terrain formed by stacking and arranging a plurality of Agaricus bisporus cultivation shelves on the ground. The roads under this terrain are always straight and the intersections are right-angled intersections.

4. A method for utilizing the robot navigation system for Agaricus bisporus cultivation shelves according to claim 1, characterized in that: The steps include: Step 1: The robot is powered on and the controller initializes the system. Step 2: The controller turns on the laser rangefinder array and accelerometer to collect environmental data; Step 3: As the robot moves forward, the lateral deviation correction group sensor of the laser rangefinder array emits laser light to measure the distance to the planting slots of the Agaricus bisporus cultivation rack, calculates the robot's posture, and makes real-time deviation corrections to ensure straight movement. Step 4: The controller runs the intersection detection algorithm and the stop detection algorithm in real time; Step 5: After reaching the manually set end point, if the robot needs to return, the original backward direction is changed to the new forward direction, the recognition position of each sensor of the laser rangefinder array is reset, the forward direction of the robot is reversed, and then the process returns to step 3 until the user actively stops the robot or the number of round trips reaches the set value.

5. A robot navigation method for Agaricus bisporus cultivation shelves according to claim 4, characterized in that: Step 4 specifically includes: Step 4.1: The controller runs the parking detection algorithm. If the parking condition is met, the robot will pause and wait for the parking time to end before resuming operation. Step 4.2: The controller runs the intersection detection algorithm, detects the road type ahead, and selects a road to move forward according to the preset road selection strategy.

6. A robot navigation method for Agaricus bisporus cultivation shelves according to claim 5, characterized in that: The road types include straight roads, crossroads, left L-shaped intersections, right L-shaped intersections, T-shaped intersections, left-turn intersections, and right-turn intersections.

7. The robot navigation method for Agaricus bisporus cultivation shelves according to claim 5, characterized in that: The road selection strategy includes setting priorities for right turn, left turn and going straight.

8. The robot navigation method for Agaricus bisporus cultivation shelves according to claim 4, characterized in that: The intersection detection algorithm includes: (1) Straight path determination: When the measurement value of the first sensor (101) exceeds 1.5W corridor , W corridor is the standard width of the corridor, and the fluctuation range of the measured values ​​of the third sensor (103) and the fourth sensor (104) is ≤0.1W corridor , it is determined to enter the straight road; (2) Intersection determination: When the measurement value of the first sensor (101) continues to decrease, and the measurement values ​​of the third sensor (103) and the fourth sensor (104) are both greater than 1.5W corridor The controller determines that the vehicle has entered an intersection and pushes the intersection record onto the stack; (3) Left L-shaped intersection judgment: When the first sensor (101) exceeds 1.5W corridor , the measurement value of the third sensor (103) suddenly increases to ≥2W corridor , and the fluctuation range of the fourth sensor (104) measurement value is ≤0.1W corridor , the controller determines that it has entered a left L-shaped intersection and pushes the intersection record into the stack; (4) Right L-shaped intersection judgment: When the first sensor (101) exceeds 1.5W corridor , the fourth sensor (104) measurement value suddenly increases to ≥2W corridor , and the third sensor (103) measurement value fluctuation range ≤ 0.1W corridor , the controller determines it as a right L-shaped intersection and pushes the intersection record into the stack; (5) T-junction determination: When the measurement value of the first sensor (101) continues to decrease, and the measurement values ​​of the third sensor (103) and the fourth sensor (104) are both greater than 1.5W corridor , the controller determines it as a T-junction and pushes the record of this intersection into the stack; (6) Left turn intersection judgment: When the measurement value of the first sensor (101) continues to decrease, and the measurement value of the third sensor (103) suddenly increases to ≥2W corridor , and the fluctuation range of the fourth sensor (104) measurement value is ≤0.1W corridor , the controller determines it as a left turn intersection and pushes the intersection record into the stack; (7) Right turn intersection judgment: When the measurement value of the first sensor (101) continues to decrease, and the measurement value of the fourth sensor (104) suddenly increases to ≥2W corridor , while the third sensor (103) measurement value fluctuation range ≤0.1W corridor , the controller determines it as a right turn intersection and pushes the intersection record into the stack; (8) Blockage and fallback judgment: When the measurement value of the first sensor (101) is ≤2W corridor , and the fluctuation range of the measurement values ​​of the third sensor (103) and the fourth sensor (104) is ≤0.1W corridor , stop and wait for 10 seconds. During the waiting period, if the measurement value of the first sensor (101) exceeds 1.5W corridor , the robot continues to move forward. If the waiting time ends and the measurement value of the first sensor (101) does not change, the controller determines that the current road branch has been completed, marks this branch, returns to the previous intersection, and selects other road branches to move forward. If there is no other branch to choose from at the current intersection, the controller determines that this intersection has been completely traversed, pops the record of this intersection out of the stack, and returns to the previous intersection of this intersection to continue judging until the stack is empty or the end point is reached.

9. The robot navigation method for Agaricus bisporus cultivation shelves according to claim 4, characterized in that: The parking algorithm includes: (1) The controller collects the pulse signal of the shelf laser sweeping across the support column through the seventh sensor (107) and the eighth sensor (108). When the controller detects that the distance value suddenly changes to D leg ≤W corridor When the duration Δt is within the set range, it is defined as passing through a support column; D leg W is the distance from the laser emitted by the sensor to the laser returned by the support column. corridor is the standard width of the corridor; (2) Establish a support column pulse counter, the count value is C leg , the initial value is the number of remaining support columns N before the target parking position target ; (3) Every time it passes an effective support column, C leg =C leg -1, when C leg = 0, the deceleration and parking program is triggered, and the control law v(t) is: v(t)=v0·e -λt v0 is the maximum speed, λ is the deceleration coefficient, and t is the time; (4) After parking, calibrate the lateral distance from the lateral deviation correction group sensor to the planting groove to satisfy ∣D side -D set ∣ <u, then lock the wheels. Dside represents the distance between the lateral deviation correction group sensor and the planting groove, Dset is the set threshold, and u is the set target difference.

10. The robot navigation method for Agaricus bisporus cultivation shelves according to claim 4, characterized in that: In step 3, the third sensor to the sixth sensor measure the real-time distance between each sensor and the side wall of the planting trough, and generate a correction instruction through the PID algorithm.

Images