A giant panda shelter intelligent access control system and method

Abstract

The application discloses a giant panda zoo intelligent access control system and method, belonging to the field of zoo facilities and intelligent control technology; a multi-source sensing module can read the panda individual ID and call the body size parameters and the rest file, and can identify the panda posture and the in-zone / out-of-zone state in real time on both sides of the door body and the door closing influence area; a determination control module executes door opening and closing according to the area state or manual instruction, and only executes the anti-pinch door closing when the panda is in the in-zone state; during the anti-pinch process, a safe height determination unit adopts three-stage variable speed descending based on the individual body size and posture; combined with the tension, displacement, pressure and infrared grating data, passive resistance and active force resistance are accurately distinguished, and a differentiated response strategy is executed; the system effectively solves the problems of single anti-pinch logic and rigid safety threshold of the traditional access control, and significantly improves the safety, adaptability and automation level of the giant panda zoo management by using a lightweight determination method.

Description

Technical Field

[0001] This invention relates to the field of zoo enclosure facilities and intelligent control technology, specifically to an intelligent access control system and method for giant panda enclosures. Background Technology

[0002] As a Class I protected animal in China, the safety, comfort, and automation level of the panda enclosures directly affect the pandas' welfare and breeding efficiency. Currently, most panda breeding bases and zoos in China use traditional manual sliding doors or simple electric lifting doors for access control, which have the following core technological shortcomings: First, the anti-pinch control logic is simplistic and lacks scene differentiation capabilities. Most existing electric gates use single current detection or infrared beam detection to achieve obstacle detection and rebound, which cannot accurately distinguish whether the pinched object is a rigid debris or a panda limb. This either leads to frequent false triggers resulting in extremely low closing efficiency, or insufficient sensitivity poses a risk of pinching injury, failing to balance safety and operational efficiency.

[0003] Second, the safety thresholds are rigid and cannot adapt to individual and posture differences. Existing solutions generally use a uniform and fixed safety height threshold, which can only be set according to the most conservative crawling posture in order to cover all risk scenarios. This results in an overly conservative approach for adult individuals in a standing posture, leading to frequent unnecessary full-opening and back-off during the closing process, which is inefficient. At the same time, it cannot adapt to the size differences of adults, sub-adults, and cubs, and the level of precision is insufficient.

[0004] Third, it lacks the ability to determine the direction of passage and has a low degree of automation. The existing access control system cannot autonomously determine the panda's passage intention and the area it should go to. It either requires manual operation throughout the process or mechanically executes timed door opening and closing, which cannot adapt to individual differences in daily routines. During the closing process, it cannot distinguish whether the panda is in the target control area, which easily leads to the risk of accidental door closing, injury, or escape. Summary of the Invention

[0005] To address the shortcomings of the existing technologies, this invention aims to provide an intelligent access control system and method for giant panda enclosures, solving the problems of high risk of pinching injuries and low closing efficiency caused by the simple anti-pinch control logic and rigid safety thresholds in existing giant panda enclosure access control systems.

[0006] To solve the above problems, the present invention adopts the following technical solution: A smart access control system for a giant panda enclosure includes a composite door, a multi-source sensing module, and a judgment and control module; the composite door includes a movable door driven by a lifting transmission mechanism; the multi-source sensing module includes an individual identification unit and a two-way early warning unit. The individual identification unit is used to read the individual ID of the panda and retrieve the corresponding individual's body size parameters and daily routine records; The two-way early warning unit is used to determine whether a panda is in the area it should be in or not in the area affected by the closed door, based on the panda's daily routine records. The determination and control module includes a door opening / closing command execution unit; The door opening / closing command execution unit is used to execute door opening or closing based on whether the panda in the door closing influence area is in the required or unrequired area, or based on manual command. When executing the door closing command, if the panda in the door closing influence area is only in the required area, then the anti-pinch door closing is executed; if the panda in the door closing influence area is in the unrequired area, then the door opening command is executed.

[0007] The beneficial effects of this invention are as follows: First, by reading the panda's ID through the individual identification unit and retrieving its body size parameters and daily routine records, combined with the two-way early warning unit, the system can determine in real time whether the panda is in the area it should be in or should not be in the area affected by the door closing. This enables the access control logic to have individualized and spatiotemporal adaptability capabilities, and only executes anti-pinch door closing when the panda is in the area it should be in, thus avoiding the risks of accidental door closing, pinching, and escape, and significantly improving the level of automated management.

[0008] Second, the safety height determination unit determines the static and dynamic safety heights based on the individual's body size and real-time posture, dividing the door's travel into three zones: high, middle, and low. Correspondingly, it employs a first, second, and third decreasing speed for variable-speed descent. This graded speed adjustment strategy avoids overly conservative regression to a rigid threshold for standing adult individuals while ensuring a safety margin in a crawling posture, significantly improving door-closing efficiency while guaranteeing safety.

[0009] Third, the state determination unit and the scene analysis unit integrate multi-source data such as wire rope tension, position encoder, door bottom pressure array and infrared grating to accurately distinguish between passive obstruction and active force obstruction, and execute differentiated responses for different scenarios. The refined differentiation strategy effectively avoids the problem of frequent false triggering or insufficient sensitivity of traditional single current detection, and achieves a balance between safety and operating efficiency.

[0010] Fourth, the escape detection unit uses the direction of change in infrared grating data or RFID reading sequence to determine the direction of the panda's passage, automatically identifying escape and compliant passage, further reducing the need for manual intervention and improving the level of intelligence and unmanned operation of the animal enclosure management. Attached Figure Description

[0011] Figure 1 This is a schematic diagram of a composite door structure.

[0012] Figure 2 A flowchart for dividing the passive obstruction state into two categories: holding debris and holding panda limbs.

[0013] Figure 3A flowchart for classifying the active force application encountering resistance state into normal mechanical disturbance and panda lifting door.

[0014] Figure 4 This is a flowchart for correcting the classification of passive obstacle detection state and active force-applied obstacle detection state using infrared grating data of the door frame. Detailed Implementation

[0015] The present invention will be further described in detail below with reference to specific embodiments.

[0016] It should be noted that these embodiments are only used to illustrate the present invention and are not intended to limit the present invention. Simple improvements to the method under the premise of the present invention are all within the scope of protection claimed by the present invention.

[0017] Example 1 See Figure 1 The composite door is divided into an upper shielding area, a middle containment area, and a lower passage area along its height. The upper shielding area is used to shield the space above the doorway to prevent giant pandas from climbing over it. The middle containment area is used to house the movable door. The lower passage area is a movable door driven by a lifting transmission mechanism, used for opening, closing, and gap control of the lower passage area. The composite door also includes a door frame, and the left column of the door frame is connected to the pre-embedded steel base in the animal enclosure wall through two sets of heavy-duty load-bearing hinges, allowing the entire door frame to be rotated around the left column.

[0018] The door frame is welded together from left column 1, right column 2, upper beam 3, and lower beam 4. The whole structure is a rectangular frame, and all main materials are made of 304 stainless steel to meet the requirements of the humid and corrosive environment of the animal shed.

[0019] The uprights are made of 60mm×60mm×3mm stainless steel square tubing to ensure sufficient structural strength to withstand panda impacts. The left upright is connected to the upper and lower crossbeams via two sets of heavy-duty load-bearing hinges, allowing the entire door frame to rotate 0-90° around the left upright, meeting the needs for wide-open passage. An electric door opener 8 can also be installed on the upper load-bearing hinge, enabling the composite door to automatically open and close based on command rotation. The right upright is equipped with an electromechanical lock body 9, which can cooperate with the lock seat on the animal enclosure wall to lock and fix the entire door frame.

[0020] The upper crossbeam is made of 60mm×60mm×3mm stainless steel square tubing, with a length of 1500mm (fitting the width of a standard animal shed doorway). Both ends are welded and fixed to the top of the left and right columns, serving as the installation base for the transmission system and the protective shell.

[0021] The lower crossbeam is made of 60mm×60mm×3mm stainless steel square tubing, with both ends welded and fixed to the bottom of the left and right columns. The bottom is flush with the ground, serving as the limiting base for the sliding door to close. It is also equipped with positioning pin holes and mechanical locks to ensure the positional accuracy and locking reliability of the sliding door after it is closed.

[0022] The lower passage area 7 has a height range of 700mm-1000mm and serves as the operating range for the lower movable door. The movable door rises and falls within this range to control the opening, closing, and clearance of the lower part of the doorway, and supports suspension and locking at any position. The width and height range of the upper obstruction area 5 and the middle accommodating area 6 are set based on the actual door frame height and the specific height of the lower passage area.

[0023] When the movable door is fully lowered, its bottom fits against the lower crossbeam, completely sealing the doorway. When fully raised, its bottom is flush with the top of the central enclosure, and the maximum passage height at the bottom of the doorway is sufficient for adult giant pandas to pass through normally. Simultaneously, the movable door can be suspended and locked at any height, leaving a certain gap to prevent it from crushing bamboo, facilitate cleaning, and accommodate scenarios where pandas lift the door.

[0024] The surface of the composite door is a composite layer, with a focus on the central storage area and / or the lower passage area. The outer layer of the composite layer is made of silicone rubber, the middle layer is made of aramid fiber compound rubber, and the base layer is stainless steel.

[0025] More specifically: the surface layer is made of 8mm thick food-grade silicone rubber with a Shore hardness controlled at 25±5, close to the natural hardness of a giant panda's paw pads. When a panda's front paws rest on the door, the hardness of the paw pads and the surface material is similar, making it impossible to form a rigid fulcrum for gripping feedback, making it difficult to exert force for climbing, naturally reducing the desire to climb and explore. When biting, the teeth contact the flexible surface, preventing them from hitting the rigid substrate and avoiding tooth breakage; during impacts, the impact force on the paws and nose is cushioned, reducing stress and physical damage.

[0026] The middle layer is made of 12mm thick aramid fiber-blended rubber material with a damping coefficient of 0.65, tensile strength ≥20MPa, and tear strength ≥50kN / m. For example, against the peak impact force and bite force of an adult giant panda of 1800N, this layer can dissipate more than 60% of the impact energy, preventing the impact force from being directly transmitted to the base layer and causing deformation of the door structure, and preventing rigid reaction force rebound damage to the panda's jawbone and palm joints. The incorporated aramid fibers significantly improve the material's tear and puncture resistance, resisting panda bites and claw cuts, extending its service life, and preventing debris from being accidentally ingested by the panda.

[0027] The base layer consists of the existing stainless steel door frame, which serves as the basic load-bearing structure and does not come into direct contact with animals.

[0028] The movable door of the composite door is driven by a lifting transmission mechanism. The lifting transmission mechanism includes a drive motor, a wire rope and pulley block, and a lifting guide rail.

[0029] The drive motor is a 24V DC geared motor with a rated power of 300W, a reduction ratio of 1:180, and a rated output torque of 120N·m. It has a power-off brake interface and is installed on the upper crossbeam or an external motor mounting bracket. The motor output shaft is connected to a winch drum with a diameter of 100mm. The wire rope is wound and unwound by forward and reverse rotation. Variable damping countermeasures are achieved through a closed-loop current control of the drive motor: the controller collects real-time data on the wire rope tension changes caused by the panda door lifting, dynamically adjusts the motor winding current, and outputs a damping torque that is opposite in direction and adjustable in magnitude to the door lifting force, limiting the door displacement within a preset safety distance until the door lift disappears.

[0030] The steel wire rope and pulley system uses two φ6mm 304 stainless steel wire ropes, which are connected to the left and right suspension points at the top of the sliding door, respectively. The wire ropes go up and around the fixed pulley at the top, and finally are wound and fixed on the winch drum. The two wire ropes are raised and lowered synchronously to ensure that the door moves horizontally with an inclination of ≤1mm. The fixed pulley seat has a built-in S-type tension sensor to detect the tension of the steel wire rope in real time.

[0031] The lifting guide rails consist of a high-strength aluminum alloy guide rail installed on the inner side of each of the left and right columns in the central receiving area. The surface is treated with hard anodizing. Wear-resistant nylon sliders are embedded in the guide rail grooves, and these sliders are bolted to the sides of the sliding door. The guide rails feature a C-shaped cross-section design, ensuring both lifting and guiding accuracy while limiting the forward and backward swaying of the sliding door and resisting the lateral impact of the door, with a resistance to lateral impact force ≥2000N.

[0032] The sliding door adopts a dual locking scheme of electronic brake and mechanical one-way anti-reverse, which prevents the door from accidentally sliding down or being lifted up by the panda, and supports suspension and locking at any position.

[0033] The electro-electromagnetic brake is integrated into the tail of the geared motor. It is a normally closed power-off brake that releases when powered on and locks the motor output shaft when powered off. It can quickly lock the door at any position with a response time of ≤100ms and a locking torque of ≥200N・m.

[0034] The mechanical one-way check ratchet is installed at the end of the winch drum and is a pawl-type one-way mechanism. When the motor is actively controlled, the pawl can be lifted by the electromagnetic lever, allowing the drum to rotate in both directions to raise and lower the door. When the system is powered off, the pawl is released and engaged in the ratchet tooth groove, allowing the drum to rotate only in the direction of rope retraction and door raising, and prohibiting the reverse rotation of rope release and door lowering. It is unaffected by electrical control system failures.

[0035] Example 2 This embodiment provides an intelligent access control system for a giant panda enclosure, including the aforementioned composite door, a multi-source sensing module, and a determination and control module. The multi-source sensing module includes an individual identification unit and a two-way early warning unit; the determination and control module includes a door opening / closing command execution unit.

[0036] The individual identification unit is used to read the individual ID of the panda and retrieve the corresponding individual's body size parameters and daily routine records.

[0037] This embodiment provides a feasible configuration method for the individual identification unit: Using RFID radio frequency identification technology, based on the subcutaneous electronic chip already implanted in the giant panda, no additional invasive procedures are required on the animal.

[0038] The chip carrier is implanted under the skin of the giant panda's shoulder blade, storing a unique individual ID and a complete file corresponding to the individual management database in the park.

[0039] One RFID reader antenna is installed at a low position on each of the inner and outer sides of the left or right pillars of the doorway, facing the corresponding area. The identification range covers the entire doorway area and a 1-meter surrounding area. The reader can simultaneously read the chip IDs of multiple individuals, connecting to the park's panda individual management database. It can retrieve real-time data such as body size parameters, daily routines, and behavioral characteristics of the corresponding individual through the chip ID. The individual identification unit also determines whether the panda is currently inside or outside the door, providing location information for determining the panda's activity direction and catching escape attempts. For example, by sorting the devices that identified the chip IDs based on time, this sorting can represent the direction the panda passed through the door.

[0040] The two-way early warning unit is used to identify the panda's posture on both sides of the composite door and within the door frame's closing influence area, and to determine whether the panda is in the appropriate or inappropriate area within the closing influence area based on the panda's daily routine.

[0041] The door-closing impact area consists of two identical single-level early warning detection mechanisms set up on the inside and outside of the door. The core function is to identify the panda's posture, detect its approach, capture its movement sequence, pre-adjust control parameters, eliminate response lag, and provide data for escape determination before the panda enters the doorway.

[0042] The single-level early warning detection system consists of a row of diffuse reflection infrared grating ranging sensor arrays vertically installed on the inner and outer sides of the pillars on the left side of the doorway and on the door frame. For example, a single array has 10 measuring points, arranged from the ground to a height of 70-100cm, with a spacing of 7cm between the measuring points. The sensor detection direction is directly facing the corresponding side area, and the effective detection distance covers a range of 0.8-1.5m. Installation on one side is sufficient for depth detection, eliminating the need for additional equipment on the opposite wall, saving space in the animal enclosure passageway, and reducing the risk of damage from panda collisions.

[0043] This embodiment provides a feasible method for recognizing panda postures: By extracting four purely geometric features of the occlusion contour solely from the ranging results at different heights of the infrared grating, five pose types can be distinguished. This process does not involve dynamic parameters such as time or velocity; it relies solely on the spatial contour determination of the current frame. Bipedal standing posture: maximum detection height ≥ 60cm, width-to-height ratio < 0.6, centroid height ≥ 35cm, longitudinal continuity > 80%; Four-legged standing / walking posture: maximum detection height 45-65cm, width-to-height ratio 0.6-1.0, centroid height 25-40cm; Sitting / semi-squatting posture: maximum detection height 30-50cm, width-to-height ratio 1.0-1.5, centroid height 15-30cm; Quadrupedal crouching posture: maximum detection height ≤ 25cm, width-to-height ratio > 2.0, centroid height ≤ 12cm, longitudinal continuity < 30%; Side-lying / reclining posture: maximum detection height 20-35cm, width-to-height ratio >1.5, lateral extension length on one side >60cm.

[0044] The specific judgment value can also be adapted and adjusted by retrieving the corresponding individual's body size parameters based on the individual panda's individual ID read by the individual identification unit, which can more accurately ensure posture recognition.

[0045] Based on the pandas' daily routine records, it is determined whether the pandas are in the designated or undesignated areas within the gate-closing influence zone. That is, the pandas' daily routine records show whether the pandas are inside or outside the enclosure at different times of the day in history. The pandas' current position within the gate-closing influence zone is compared with the corresponding position recorded in the daily routine records. If they match, the pandas are in the designated area; if they do not match, the pandas are in the undesignated area.

[0046] The door opening / closing command execution unit is used to execute door opening or closing based on whether the panda in the door closing influence area is in the required or unrequired area, or based on manual instructions. When executing the door closing command, if the panda in the door closing influence area is only in the required area, the door will be closed in an anti-pinch manner; if the panda in the door closing influence area is in the unrequired area, the door will be opened.

[0047] The execution logic for the anti-pinch door closing mechanism includes: The sensor detection unit of the multi-source sensing module is used to acquire tension data from the wire rope tension sensor in the lifting transmission mechanism, displacement data from the movable door position encoder, and pressure data from the distributed pressure array at the bottom edge of the movable door.

[0048] The wire rope tension sensor can be an S-type tension / compression sensor with a range of 0–2000 N and an overall accuracy of 0.05% FS. It is installed on the top fixed ends of the left and right wire ropes and integrated with the fixed pulley seat.

[0049] The movable door position encoder can be selected as a 17-bit resolution multi-turn absolute encoder, which is installed on the shaft end of the hoisting drum and rotates coaxially with the drum to detect the precise position and displacement status of the door in real time.

[0050] The distributed pressure array at the bottom edge of the movable door can be equipped with a thin-film pressure sensor (FSR) with a range of 0–500 N. These sensors are evenly arranged along the bottom of the movable door, for example, 30 sensor points with a lateral spacing of 5 cm, covering the entire width of the door, with the sensing surface facing the ground. This allows the sensor to collect the spatial distribution characteristics of the pressure of objects in contact with the bottom of the door, distinguishing between point-like rigid objects and cluster-like flexible objects.

[0051] All sensor data are timestamped and aligned through the same programmable logic controller (PLC) synchronous acquisition module, with a time synchronization accuracy of ≤1ms, ensuring that features of different dimensions correspond completely on the time axis and avoiding judgment errors due to timing misalignment.

[0052] After the raw data is collected, it undergoes standardized preprocessing before being judged and used. Filtering and noise reduction: A 5th-order Butterworth low-pass filter is used for tension and pressure to filter out high-frequency noise caused by motor operation and mechanical resonance, while retaining the true force and vibration characteristics; Data smoothing: The position encoder data is processed by moving average with a window size of 10 sampling points to eliminate the small jitter of pulse counting and ensure the reliability of micro-motion detection; Feature extraction: Connected component extraction is performed on infrared grating and ranging array data to remove isolated noise points and calculate structured features such as height, width, and centroid position of continuously occluded areas.

[0053] The safety height determination unit of the judgment control module is used to determine the static safety height and dynamic safety height based on the body size parameters of the corresponding individual and the panda's posture. The static safety height is the maximum distance between the panda and the ground that prevents the panda from passing through the movable door, and the dynamic safety height is the height value at which the panda will not be crushed in the posture within the door's influence area. Based on the static safety height and dynamic safety height, the movable door's activity height is divided into a high-level zone, a middle-level zone, and a low-level zone.

[0054] As one possible implementation method, the static safety height can be determined based on the maximum vertical height at which a panda cannot directly squeeze through a door gap when it is in a crouching position on all fours, and a certain safety margin (such as 10%-15% of the maximum vertical height).

[0055] As one possible implementation method, dynamic safety height is determined for different pandas and different postures. For example, panda A and panda B have different body sizes, and they also differ in the same posture. Dynamic safety height can be set / saved in the individual's body size parameters for different postures.

[0056] The high-level zone is the area greater than or equal to the dynamic safety height, the mid-level zone is the area between the dynamic safety height and the static safety height, and the low-level zone is the area less than or equal to the static safety height.

[0057] When the door opening / closing command execution unit performs anti-pinch closing and the moving door is in normal operating condition, the door descent speed is set to the first descent speed in the high position zone, the door descent speed is set to the second descent speed which is less than the first descent speed in the middle position zone, and the door descent speed is set to the third descent speed which is less than the second descent speed in the low position zone.

[0058] When implementing the anti-pinch door closing mechanism, if a panda is detected within the door's closing influence area, the static and dynamic safety heights are pre-assessed or determined. The descent speed of the sliding door is then controlled in advance to prevent excessive speed and avoid the panda rushing through the door and causing injury. The dynamic safety height changes according to the panda's real-time posture, allowing for flexible adjustment of the sliding door's descent speed.

[0059] This embodiment also provides a feasible configuration method for the determination control module, which further includes a state determination unit and a scene analysis unit.

[0060] The status determination unit is used to determine whether the movable door is in normal operation, passive obstruction, or active force obstruction state when performing anti-pinch closing, based on the tension data from the wire rope tension sensor and the displacement data from the movable door position encoder.

[0061] When the door is suspended at rest, the tension of the steel wire rope is equal to the weight of the door (rated self-weight tension, defined as 100% of the baseline value). Whether the door is passively obstructed or actively lifted, the essence is that an external force lifts the door upward and shares the weight of the door, so the tension of the steel wire rope will decrease. The core difference between the two is the direction and trend of displacement: passive obstruction only hinders the fall and does not have a continuous upward displacement, while active force will produce a continuous upward displacement trend.

[0062] If the movement state of the sliding door is consistent with the control command, and the tension is stable within the first range of the rated tension of the sliding door's own weight (such as ±1%), it is determined that the sliding door is in normal operation, and the sliding door descends normally according to the speed of each zone.

[0063] If the tension amplitude reaches the second range of the rated tension of the sliding door's own weight (such as greater than 1%, 5%, etc.), and the displacement data of the sliding door position encoder shows no upward displacement trend, it is determined that the sliding door is in a passive obstruction state. This could be due to obstacles below, such as bamboo or panda limbs, during the descent.

[0064] If the tension amplitude reaches the third range of the rated tension of the sliding door's own weight (e.g., higher than 30%; the panda's lifting force offsets part of the door's own weight, reducing the tension of the steel wire rope; the greater the lifting force, the greater the tension drop), and the displacement data of the sliding door's position encoder shows an upward displacement trend and the cumulative displacement exceeds the mechanical clearance threshold, it is determined that the sliding door is in a state of active force application encountering resistance. Examples include the panda pulling at the door while it is stationary and hovering, the panda reaching up to lift the door during the descent, and the panda using its body to lift the door from below.

[0065] The scene analysis unit is used to classify the passive obstruction state into clamping debris and clamping panda limbs based on tension data and distributed pressure array data at the bottom edge of the movable door. It is also used to classify the active force obstruction state into normal mechanical disturbance and panda lifting the door based on tension data and displacement data from the movable door position encoder. Furthermore, it uses infrared grating data from the door frame to correct the classification of the passive obstruction state and the active force obstruction state.

[0066] See Figure 2 This embodiment provides a feasible way to divide the passive obstruction state into clamping debris and clamping panda limbs: S101. Analyze the drop tension amplitude change characteristics, final drop tension amplitude characteristics, and post-drop tension amplitude fluctuation characteristics of the tension amplitude data. Analyze the membership degree of the tension amplitude data to determine whether it belongs to the clamped debris or the clamped panda limbs, and perform a weighted summation (or an equal weight value, adjusted based on the accuracy of the analysis) to obtain the initial tension characteristic membership degree.

[0067] This embodiment provides one possible implementation of S101. The dimensions and calculation of the initial tension feature membership degree are shown in Table 1.

[0068] Table 1. Initial Tension Feature Membership Dimension Table

[0069] The values ​​in the table, such as "final drop tension amplitude ≥ 10%" and "tension amplitude fluctuates by more than 5% of the drop tension amplitude", can be adjusted according to the accuracy of the analysis, and there are no restrictions here.

[0070] S102. Analyze the total force area characteristics, pressure distribution gradient characteristics, and force lateral distribution characteristics of the distributed pressure array data at the bottom edge of the active door. Perform a weighted summation of the membership degrees of the objects being clamped and the panda limbs being clamped (or the weight values ​​can be evenly distributed and adjusted based on the accuracy of the analysis) to obtain the initial pressure characteristic membership degrees.

[0071] This embodiment provides one possible implementation of S102. The dimensions and calculation of the initial pressure characteristic membership degree are shown in Table 2.

[0072] Table 2 Initial Pressure Feature Membership Dimension Table

[0073] S103. The initial membership degree of the tension feature and the initial membership degree of the pressure feature are weighted and summed to obtain the initial classification type membership degree of the clamped debris and the clamped panda limbs (the weight values ​​can be evenly distributed or adjusted based on the accuracy of the analysis).

[0074] See Figure 3 This embodiment provides a feasible implementation method for classifying the active force application resistance state into normal mechanical disturbance and panda-like door-lifting: S111. Analyze the membership degree of the cumulative changes in tension and displacement to normal mechanical disturbance and panda door lifting, and perform weighted summation (the weight values ​​can be evenly distributed or adjusted based on the accuracy of the analysis) to obtain the initial tension-displacement change characteristic membership degree.

[0075] This embodiment provides a feasible method for calculating the membership degree of initial tension-displacement change characteristics: The dimensions of the membership degree of the initial tension-displacement change characteristics are shown in Table 3.

[0076] Table 3 Membership Dimension Table of Initial Tension-Displacement Change Characteristics

[0077] S112. Analyze the ratio characteristics of cumulative displacement amplitude to mechanical clearance threshold and the magnitude characteristics of cumulative tension decrease amplitude to rated self-weight tension. Analyze the membership degree of normal mechanical disturbance and panda door lifting, and sum them by weight to obtain the membership degree of initial tension-displacement amplitude characteristics.

[0078] This embodiment provides a feasible method for calculating the membership degree of the initial tension-displacement amplitude characteristics: The dimensions and calculation of the membership degree of the initial tension-displacement amplitude feature are shown in Table 4.

[0079] Table 4 Membership Dimension Table of Initial Tension-Displacement Amplitude Characteristics

[0080] S113. The initial classification type membership degree of normal mechanical disturbance and panda door lifting is obtained by weighted summation of the initial tension-displacement change characteristic membership degree and the initial tension-displacement amplitude characteristic membership degree. The weights of the initial tension-displacement change characteristic membership degree and the initial tension-displacement amplitude characteristic membership degree can be 0.7 and 0.3 respectively, or adjusted based on the accuracy of the analysis.

[0081] See Figure 4 This embodiment provides a feasible method for correcting the classification of passive obstruction states and active force-applied obstruction states using infrared grating data of the door frame, including: S121. For the classification of passive obstruction state, the occlusion height range feature and contour continuity feature of the infrared grating data of the door frame are analyzed to determine the membership degree of the clamped debris and the clamped panda limbs, and the weighted sum is performed to obtain the passive obstruction state classification correction membership degree (see Table 5). The passive obstruction state classification correction membership degree is weighted and summed with the initial classification type membership degree of clamped debris and clamped panda limbs to obtain the final passive obstruction state classification membership degree. The category with the final passive obstruction state classification membership degree greater than the corresponding classification category threshold is selected as the classification category of passive obstruction state.

[0082] Table 5. Modified Membership Feature Table for Passive Obstacle Encounter State Classification

[0083] S122. For the classification of active force-induced obstruction state, the membership degree of the occlusion height range feature and contour continuity feature of the infrared grating data of the door frame belonging to normal mechanical disturbance and panda door lifting is analyzed and weighted and summed (Table 6) to obtain the active obstruction state classification correction membership degree; the active obstruction state classification correction membership degree is weighted and summed with the initial classification type membership degree of normal mechanical disturbance and panda door lifting to obtain the final active force-induced obstruction state classification membership degree; the category with the final active force-induced obstruction state classification membership degree greater than the corresponding classification category threshold is selected as the classification category of active force-induced obstruction state.

[0084] Table 6. Calculation of Dimensions for Corrected Membership Degrees in Active Obstacle Encounter State Classification

[0085] The weight ratio for adjusting membership degrees is 20-40%, used to adjust the membership degrees of the initial classification types. The classification threshold can be set to 90% or 95%, depending on the accuracy of the analysis.

[0086] When the passive obstruction state is classified as clamping debris, if the movable door is in the high or middle zone, the lifting transmission mechanism drives the movable door to hover and move upward to leave an active area, and reports that there is debris interfering with closing the door and there is a risk of escape. If the movable door is in the low zone, the lifting transmission mechanism drives the movable door to hover and move upward to leave an active area, and ensures that the height of the movable door is not higher than the static safety height. When the passive obstruction state is classified as clamping panda limbs, if the movable door is in the middle or low position zone, the lifting transmission mechanism drives the movable door to hover and move upward to leave a safe area. After a preset time interval, the state of the movable door is judged again. If the movable door is in the normal operating state, it continues to descend; if the movable door is in the high position zone, it continues to descend at a speed less than the first descent speed. When the active force encounters resistance, the classification is normal mechanical disturbance, and the movable gate descends based on being in normal operating condition; When the active force application encounters resistance, the classification is "panda lifting door". If the movable door is in the high position zone, the lifting transmission mechanism activates variable damping countermeasure according to the force exerted by the panda lifting door, controlling the movable door to rise or fall within the preset safety distance within the height of the panda lifting door until the panda lifting door disappears, then continues to descend. If the movable door is in the middle or low position zone, the lifting transmission mechanism drives the movable door to hover and moves upward to leave a safe area, and judges the status of the movable door again after a preset time interval. If the movable door is in normal operating condition, then continues to descend.

[0087] In one configurable manner of the determination control module in this embodiment, the determination control module further includes an escape determination unit. The escape determination unit is used to acquire the panda's individual ID and the direction of change of infrared grating data within the area affected by the closed door when the active door is not closed. If the direction of change of infrared grating data is from the panda's intended area to its unintended area, the panda is determined to have escaped; if the direction of change of infrared grating data is from the unintended area to the intended area, the panda is determined to have compliant passage. Alternatively, the devices whose chip IDs have been identified can be sorted chronologically based on the identified chip IDs. This sorting can represent the direction the panda passed through the door. If the direction the panda passed through the door is from its intended area to its unintended area, the panda is determined to have escaped; otherwise, it is determined to have compliant passage.

[0088] Example 3 Based on the intelligent access control system for giant panda enclosures provided in Embodiment 2, a method is designed to read the individual ID of a panda and retrieve the corresponding individual's body size parameters and daily routine records. Identify panda postures located on both sides of a composite door and within the door frame's closing influence area, and determine whether a panda is in the appropriate or inappropriate area within the closing influence area based on its daily routine records; The door is closed or opened based on whether the pandas within the affected area are in the designated or undesignated area, or based on manual instructions. When the door is closed, if the pandas within the affected area are only in the designated area, an anti-pinch door is closed; if the pandas within the affected area are in the undesignated area, an open door is opened.

[0089] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described with reference to preferred embodiments, those skilled in the art should understand that various changes in form and detail can be made without departing from the spirit and scope of the invention as defined in the appended claims.

Claims

1. A smart access control system for a giant panda enclosure, comprising a composite door, a multi-source sensing module, and a judgment and control module; the composite door includes a movable door driven by a lifting transmission mechanism; characterized in that, The multi-source sensing module includes an individual identification unit and a two-way early warning unit; The individual identification unit is used to read the individual ID of the panda and retrieve the corresponding individual's body size parameters and daily routine records; The two-way early warning unit is used to determine whether a panda is in the area it should be in or not in the area affected by the closed door, based on the panda's daily routine records. The determination and control module includes a door opening / closing command execution unit; The door opening / closing command execution unit is used to execute door opening or closing based on whether the panda in the door closing influence area is in the required or unrequired area, or based on manual command. When executing the door closing command, if the panda in the door closing influence area is only in the required area, then the anti-pinch door closing is executed; if the panda in the door closing influence area is in the unrequired area, then the door opening command is executed.

2. The intelligent access control system for giant panda enclosures according to claim 1, characterized in that, The multi-source sensing module also includes a sensor detection unit, which is used to acquire tension data from the wire rope tension sensor in the lifting transmission mechanism, displacement data from the movable door position encoder, and pressure data from the distributed pressure array at the bottom edge of the movable door. The judgment and control module also includes a safety height determination unit, which is used to determine the static safety height and dynamic safety height based on the body size parameters of the corresponding individual and the panda's posture located on both sides of the composite door and within the door frame's closing influence area identified by the two-way early warning unit. The static safety height is the maximum distance between the panda and the ground that prevents the panda from passing through the movable door, and the dynamic safety height is the height at which the panda will not be crushed in the closing influence area. Based on the static and dynamic safety heights, the movable door's height is divided into a high zone, a middle zone, and a low zone. When the door opening / closing command execution unit performs anti-pinch closing and the moving door is in normal operating condition, the moving door descent speed is set to a first descent speed in the high position zone, a second descent speed (less than the first descent speed) in the middle position zone, and a third descent speed (less than the second descent speed) in the low position zone.

3. The intelligent access control system for giant panda enclosures according to claim 2, characterized in that, The determination and control module also includes a state determination unit and a scene analysis unit; The state determination unit is used to determine whether the movable door is in normal operation, passive obstruction, or active force obstruction state based on the tension data of the wire rope tension sensor and the displacement data of the movable door position encoder when performing anti-pinch closing. If the movement state of the movable door is consistent with the control command and the tension is stable within the first range of the rated tension value of the movable door's own weight, it is determined that the movable door is in normal operation. If the tension amplitude reaches the second range of the rated tension of the movable door's own weight, and the displacement data of the movable door position encoder shows no upward displacement trend, it is determined that the movable door is in a passive obstruction state. If the tension amplitude reaches the third range of the rated tension of the sliding door's own weight, and the displacement data of the sliding door position encoder shows an upward displacement trend and the cumulative displacement exceeds the mechanical clearance threshold, it is determined that the sliding door is in a state of active force application encountering resistance. The scenario analysis unit is used to classify the passive obstruction state into clamping debris and clamping panda limbs based on tension data and distributed pressure array data at the bottom edge of the movable door. It is also used to classify the active force obstruction state into normal mechanical disturbance and panda lifting the door based on tension data and displacement data from the movable door position encoder. Furthermore, it uses infrared grating data from the door frame to correct the classification of the passive obstruction state and the active force obstruction state.

4. The intelligent access control system for giant panda enclosures according to claim 3, characterized in that, The classification of passive obstruction states into clamping debris and clamping panda limbs includes: The membership degrees of the falling tension amplitude change characteristics, the final falling tension amplitude characteristics, and the tension amplitude fluctuation characteristics after the fall data are analyzed to determine whether they belong to the clamped debris or the clamped panda limbs, and then weighted summation is performed to obtain the initial tension characteristic membership degree. The initial pressure feature membership degree is obtained by weighted summing of the total force area characteristics, pressure distribution gradient characteristics, and force lateral distribution characteristics of the distributed pressure array data at the bottom edge of the movable door, which belong to the clamped debris and the clamped panda limbs. The initial classification type membership degree of the clamped debris and the clamped panda limbs is obtained by weighted summation of the initial tension feature membership degree and the initial pressure feature membership degree.

5. The intelligent access control system for giant panda enclosures according to claim 4, characterized in that, The classification of the active force application encountering resistance state into normal mechanical disturbance and panda door lifting includes: The membership degrees of the cumulative changes in tension and displacement, including the synchronicity, monotonicity, and proportionality of these changes, belonging to normal mechanical disturbances and the "panda lifting door" phenomenon, are analyzed and weighted to obtain the initial tension-displacement change characteristic membership degree. The membership degree of the ratio of cumulative displacement amplitude to mechanical clearance threshold and the magnitude of cumulative tension decrease amplitude to rated self-weight tension are analyzed and weighted summed to obtain the membership degree of initial tension-displacement amplitude characteristics. The initial classification type membership degree of normal mechanical disturbance and panda door lifting is obtained by weighted summation of the membership degree of the initial tension-displacement change feature and the initial tension-displacement amplitude feature; The correction of the classification of passive obstacle detection state and active force-applied obstacle detection state using infrared grating data of the door frame includes: To classify passive obstruction states, the occlusion height range and contour continuity features of the infrared grating data of the door frame are analyzed, and the membership degrees of belonging to the categories of "clamped debris" and "clamped panda limbs" are weighted and summed to obtain the corrected membership degree for passive obstruction state classification. The corrected membership degree for passive obstruction state classification is then weighted and summed with the initial classification type membership degrees of "clamped debris" and "clamped panda limbs" to obtain the final membership degree for passive obstruction state classification. Categories with the final membership degree for passive obstruction state classification greater than the corresponding classification category threshold are selected as the classification categories for passive obstruction states. To classify the active force-induced obstruction state, the membership degrees of the occlusion height range and contour continuity features of the infrared grating data of the door frame belonging to normal mechanical disturbance and panda-style door lifting are analyzed and weighted summed to obtain the corrected membership degree for the active force-induced obstruction state classification. The corrected membership degree for the active force-induced obstruction state classification is then weighted summed with the membership degrees of the initial classification types of normal mechanical disturbance and panda-style door lifting to obtain the final membership degree for the active force-induced obstruction state classification. The categories with the final membership degree for the active force-induced obstruction state classification greater than the corresponding classification category threshold are selected as the classification categories for the active force-induced obstruction state.

6. The intelligent access control system for giant panda enclosures according to claim 5, characterized in that, When the passive obstruction state is classified as clamping debris, if the movable door is in the high or middle zone, the lifting transmission mechanism drives the movable door to hover and move upward to leave an active area, and reports that there is debris interfering with closing the door and there is a risk of escape. If the movable door is in the low zone, the lifting transmission mechanism drives the movable door to hover and move upward to leave an active area, and ensures that the height of the movable door is not higher than the static safety height. When the passive obstruction state is classified as clamping panda limbs, if the movable door is in the middle or low position zone, the lifting transmission mechanism drives the movable door to hover and move upward to leave a safe area, and judges the state of the movable door again after a preset time interval. If the movable door is in normal operation, it continues to descend. If the movable door is in the high position zone, it will continue to descend at a speed less than the first descent speed; When the active force encounters resistance, the classification is normal mechanical disturbance, and the movable gate descends based on being in normal operating condition; When the active force application encounters resistance, the classification is "panda lifting door". If the movable door is in the high position zone, the lifting transmission mechanism activates variable damping countermeasure according to the force exerted by the panda lifting door, controlling the movable door to rise or fall within the preset safety distance within the height of the panda lifting door until the panda lifting door disappears, then continues to descend. If the movable door is in the middle or low position zone, the lifting transmission mechanism drives the movable door to hover and moves upward to leave a safe area, and judges the status of the movable door again after a preset time interval. If the movable door is in normal operating condition, then continues to descend.

7. The intelligent access control system for giant panda enclosures according to claim 6, characterized in that, The determination and control module also includes an escape determination unit; the escape determination unit is used to obtain the individual ID of the panda and the direction of change of the infrared grating data of the panda in the area affected by the closed door when the activity door is not closed; if the direction of change of the infrared grating data is from the area where the panda should be to the area where it should not be, then the panda is determined to have escaped; if the direction of change of the infrared grating data is from the area where the panda should not be to the area where it should be, then the panda is determined to have passed through in compliance with regulations.

8. The intelligent access control system for giant panda enclosures according to claim 1, characterized in that, The composite door is divided into an upper shielding area, a middle accommodating area, and a lower passage area along its height. The upper shielding area is used to shield the space above the doorway to prevent giant pandas from climbing over it. The middle accommodating area is used to house the movable door. The lower passage area is a movable door driven by a lifting transmission mechanism, used for opening, closing, and gap control of the lower passage area. The composite door also includes a door frame, and the left or right column of the door frame is rotatably connected to the door frame base on the wall to achieve a rotating door opening and closing mechanism.

9. The intelligent access control system for giant panda enclosures according to claim 8, characterized in that, The surface of the central accommodating area and / or the lower passage area of ​​the composite door is a composite layer. The outer layer of the composite layer is made of silicone rubber, the middle layer is made of aramid fiber compounded rubber, and the base layer is made of stainless steel.

10. A method for intelligent access control in a giant panda enclosure, characterized in that, include: Read the individual panda's ID and retrieve the corresponding individual's body size parameters and daily routine records; Based on the pandas' daily routine records, it was determined whether the pandas in the area affected by the gate closure were in the areas where they should or should not be. The door is closed or opened based on whether the pandas within the affected area are in the designated or undesignated area, or based on manual instructions. When the door is closed, if the pandas within the affected area are only in the designated area, an anti-pinch door is closed; if the pandas within the affected area are in the undesignated area, an open door is opened.

Images