A large-stroke horizontal hydraulic upsetting mechanism with real-time pressure feedback

Abstract

The present application belongs to the technical field of metal forging, and in particular relates to a large-stroke real-time pressure feedback horizontal hydraulic upsetting mechanism, which comprises a rack, a horizontal hydraulic cylinder, an upsetting head, a pressure sensor array and a control unit. The pressure sensor array is integrated on the working surface of the upsetting head and is used for real-time monitoring of pressure distribution; the control unit establishes a geometric reference based on sensor data, calculates pressure distribution deviation and resultant force estimation through calibration and periodic acquisition, and generates an adjustment coefficient by weighting. When the adjustment coefficient continuously exceeds the threshold value and reaches the shortest residence time, a control instruction of adjustment direction and movement step length is generated to drive the hydraulic cylinder to move. Through real-time feedback and quantitative control, the present application solves the problems of uneven pressure distribution and control lag in traditional upsetting, and improves the forging precision and efficiency. The device structure is solid and suitable for large-stroke forging scenes, and the control method has traceability and reproducibility.

Description

Technical Field

[0001] This invention relates to the field of metal forging technology, and in particular to a horizontal hydraulic upsetting mechanism with large stroke and real-time pressure feedback. Background Technology

[0002] Hydraulic upsetting is a common process in metal forging, used to compress and shape workpieces. Traditional hydraulic upsetting mechanisms typically rely on operator experience or preset programs to control the stroke and pressure of the hydraulic cylinders, lacking a real-time pressure distribution monitoring and feedback mechanism. This leads to deviations in pressure distribution during long-stroke forging due to uneven workpiece material, die wear, or changes in installation posture, resulting in problems such as uneven workpiece forming, cracks, or die damage.

[0003] In existing technologies, some improved solutions use a single pressure sensor to monitor the overall pressure, but cannot capture the details of pressure distribution; other solutions indirectly infer pressure through vision or displacement sensors, but the response is lagging and is greatly affected by environmental interference. In multi-person collaboration or complex working conditions, control decisions rely on subjective judgment, which is prone to misadjustment or overadjustment. In addition, the lack of a unified geometric benchmark and quantitative criteria leads to inconsistent control calibers under different installation postures, which are difficult to reproduce and trace. Therefore, there is an urgent need to develop a hydraulic upsetting mechanism with real-time pressure feedback, adaptive adjustment and traceable control to improve forging accuracy and reliability. Summary of the Invention

[0004] Based on the technical problems existing in the prior art, this invention proposes a horizontal hydraulic upsetting mechanism with large stroke and real-time pressure feedback.

[0005] This invention proposes a horizontal hydraulic upsetting mechanism with large stroke and real-time pressure feedback, comprising a frame, a horizontally arranged hydraulic cylinder, and an upsetting head connected to the hydraulic cylinder. A pressure sensor array is mounted on the working surface of the upsetting head. A control unit is mounted on the frame and is communicatively connected to the pressure sensor array and the hydraulic cylinder. The control unit is configured to perform the following steps: establishing a unified geometric and coordinate reference; calibrating the pressure sensor array under no-load and load conditions, and preset thresholds and minimum dwell times; periodically collecting pressure data from each pressure sensor array; calculating pressure distribution deviation and resultant force estimation based on the pressure data; calculating an adjustment coefficient by weighted summation of the pressure distribution deviation and resultant force estimation, and comparing the adjustment coefficient with a preset threshold; when the adjustment coefficient continuously exceeds the preset threshold and reaches the predetermined minimum dwell time, generating a control command containing the adjustment direction and movement step size, and controlling the hydraulic cylinder to move; and executing a reset process after adjustment, and recording relevant data of this event to a non-volatile memory.

[0006] The frame is welded from high-strength steel and treated with rust prevention, providing stable support for the overall structure. The hydraulic cylinder is designed with a stroke range of 0-1000mm to meet the requirements of large-stroke forging. A pressure sensor array of 24 sensors is evenly distributed around the circumference of the upsetting head working surface, enabling full-area pressure coverage monitoring. The control unit is preferably an industrial PLC, integrating non-volatile memory and communication interface to ensure data storage security and smooth equipment linkage. In the control process, a unified geometric and coordinate reference is used with the geometric center of the upsetting head working surface as the origin, combined with the attitude sensing unit to achieve coordinate alignment, ensuring consistency in judgment under different installation postures. The calibration process includes two levels: no-load and load, which respectively determine the zero-point offset, noise level, and pressure conversion coefficient, laying the foundation for the accuracy of pressure data. The data acquisition cycle is set to 10ms, which can quickly capture dynamic changes in pressure. The recorded data covers key information such as calibration constants, threshold parameters, adjustment direction, and step size, forming a complete and traceable process data chain.

[0007] Furthermore, the upsetting head and the hydraulic cylinder are connected via a flange, with reinforcing ribs at the connection point. The flange connection offers advantages such as easy disassembly and precise positioning, facilitating subsequent equipment maintenance and upsetting head replacement. The reinforcing ribs at the connection point are made of high-strength alloy material compatible with the upsetting head and hydraulic cylinder, and are fixed by welding, significantly enhancing the structural rigidity and load-bearing capacity of the connection. During long-stroke forging, the forging force applied by the hydraulic cylinder driving the upsetting head can reach hundreds of kilonewtons. The reinforcing ribs effectively disperse stress concentration at the connection, preventing deformation and cracking under long-term high pressure, ensuring long-term stable operation of the mechanism under high-intensity, high-frequency conditions.

[0008] Furthermore, the upsetting head integrates an attitude sensing unit for outputting its attitude information. This unit outputs attitude quaternions, its core function being to align the working surface coordinates of the upsetting head to the gravity direction in world coordinates, making the horizontal direction a globally unified reference. Traditional hydraulic upsetting mechanisms often suffer from inconsistent pressure judgment benchmarks due to variations in installation posture. This attitude sensing unit, however, adjusts only the angular reference zero point through coordinate alignment, without altering the physical magnitude of the pressure, providing a unified coordinate benchmark for subsequent pressure region determination and direction adjustment decisions. Even with slight offsets in the mechanism's installation position, attitude compensation ensures accurate pressure monitoring and control, effectively solving the problem of inconsistent control parameters caused by changes in installation posture in traditional mechanisms.

[0009] Furthermore, the pressure sensor array is configured to simultaneously detect normal pressure and tangential shear force. Each pressure sensor possesses dual-dimensional detection capabilities: normal pressure data is used to calculate the core pressure value and resultant force estimate during the forging process, while tangential shear force data can assist in calculating the friction coefficient between the workpiece and the upsetting head. The combination of these two data points provides a more comprehensive and accurate reflection of pressure distribution details and contact conditions. Compared to the limitations of traditional single pressure sensors that can only monitor overall pressure, this design effectively captures local pressure deviations, such as local pressure concentration caused by uneven workpiece material or mold wear. This provides the control unit with rich decision-making information, avoiding quality defects such as uneven workpiece forming and cracks caused by incomplete pressure monitoring.

[0010] Furthermore, when the control unit calculates the pressure distribution deviation, it includes: acquiring the identification information of each triggered pressure sensor and calculating its average value; calculating the average value of the identification information of the pressure sensor corresponding to the preset preferred pressure zone; the pressure distribution deviation is the difference between the above two average values.

[0011] Furthermore, the preferred pressure zone is set as follows: the upsetting head working surface is divided radially into an inner region, a middle region, and an outer region, and the middle region is set as the preferred pressure zone.

[0012] Furthermore, the generation of control commands including adjustment direction and movement step size specifically includes: continuously monitoring the adjustment coefficient, and only confirming the need to generate control commands when it continuously exceeds a preset threshold and reaches a preset minimum dwell time; determining the adjustment direction based on the pressure distribution area on the upsetting head: if the pressure area is the inner area, the direction is indicated as backward movement; if it is the outer area, the direction is indicated as forward movement; if it is the middle area, pressure fine-tuning is prioritized; the movement step size is a discrete integer value, the magnitude of which is determined by the over-threshold amplitude and the tightness index, and an upper limit constraint is applied to this value; and generating control commands based on the adjustment direction and movement step size.

[0013] Furthermore, after the control command is issued, the system enters a latching state, during which no new commands are generated repeatedly until the operation execution result is received or the safe clearing time is reached.

[0014] Furthermore, the over-threshold amplitude is the amplitude by which the adjustment coefficient exceeds a preset threshold, and the combined force estimation is normalized to obtain the tightness index.

[0015] Compared with the prior art, the present invention provides a horizontal hydraulic upsetting mechanism with large stroke and real-time pressure feedback, which has the following advantages:

[0016] 1. Breaking through the limitations of traditional single-sensor monitoring, achieving precise capture of pressure distribution: This invention uses an array of 24 evenly distributed pressure sensors, which can simultaneously detect normal pressure and tangential shear force. Combined with 10ms high-frequency data acquisition and attitude compensation, it can locate the pressure area and distribution deviation in real time, avoiding problems such as uneven workpiece forming and cracks caused by uneven materials and mold wear. It controls the forging dimensional accuracy error within a smaller range and significantly improves the product qualification rate.

[0017] 2. Unify geometric and coordinate references to ensure control consistency under multiple working conditions: The working surface coordinates are aligned with the world coordinates through the attitude sensing unit. A standardized measurement chain is established through two-level calibration under no-load and load conditions. This solves the problems of installation attitude differences and inconsistent control calibers caused by the lack of a unified reference in the past. It makes the forging process reproducible under different batches and different installation scenarios, provides stable technical support for mass production, and reduces process debugging costs.

[0018] 3. Adaptive hierarchical adjustment mechanism to avoid the risks of misadjustment and overadjustment: The design uses a minimum dwell time of 0.5s for signal de-jittering, accurately determines the direction based on the pressure area (inner zone retreat, outer zone advance, middle zone fine adjustment), and dynamically calculates the movement step size by combining the over-threshold amplitude and the tightness index of the resultant force normalization. In addition, the command is latched for 2s after it is issued to prevent repeated triggering. This solves the problems of lag and subjectivity in traditional manual judgment or preset programs, and improves the adjustment response speed and accuracy.

[0019] 4. Full-process data traceability enhances quality control capabilities: The system writes calibration constants, threshold parameters, and adjustment events (direction, step size, pressure deviation, etc.) into non-volatile memory, forming a complete data chain from calibration to forging. This facilitates subsequent traceability of the forging process of each batch of workpieces, quickly identifies the causes of process abnormalities, and provides real data support for process optimization, helping to continuously improve production efficiency. Attached Figure Description

[0020] Figure 1 This is a schematic diagram of the overall structure of a horizontal hydraulic upsetting mechanism with large stroke and real-time pressure feedback proposed in this invention.

[0021] Figure 2 This is a schematic diagram of the back structure of the upsetting head of a horizontal hydraulic upsetting mechanism with large stroke and real-time pressure feedback proposed in this invention.

[0022] Figure 3 This is a schematic diagram of the control method for a horizontal hydraulic upsetting mechanism with large stroke and real-time pressure feedback proposed in this invention.

[0023] In the diagram: 1. Frame; 2. Hydraulic cylinder; 3. Upsetting head; 4. Pressure sensor array; 5. Control unit. Detailed Implementation

[0024] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.

[0025] In the description of this invention, it should be understood that the terms "upper", "lower", "front", "rear", "left", "right", "top", "bottom", "inner", "outer", etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are only for the convenience of describing this invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on this invention.

[0026] Example 1: A horizontal hydraulic upsetting mechanism with large stroke and real-time pressure feedback according to the present invention, such as... Figures 1 to 2 As shown, the assembly includes a frame 1, a horizontally positioned hydraulic cylinder 2, an upsetting head 3, a pressure sensor array 4, and a control unit 5. The frame 1 is welded from high-strength steel and its surface is treated with rust prevention. The hydraulic cylinder 2 is horizontally mounted on the frame 1, with a stroke range of 0-1000mm to meet the requirements of large stroke. The upsetting head 3 is connected to the hydraulic cylinder 2 via a flange, and the flange is equipped with reinforcing ribs. The pressure sensor array 4 is integrated on the working surface of the upsetting head 3, evenly distributed along the circumference, with a total number of K=24. Each sensor can simultaneously detect normal pressure and tangential shear force. The control unit 5 is mounted on the side of the hydraulic cylinder 2 and is connected to the sensors and the hydraulic cylinder 2 via cables.

[0027] Control unit 5 performs as follows Figure 3 The control method shown has the following specific steps:

[0028] Step S1: System Initialization and Calibration

[0029] After the device is powered on, it enters the initialization phase, establishing a unified geometric and coordinate reference. With the geometric center of the upsetting head 3 working surface as the origin and the working surface as the reference plane, the radial angle is defined as the unique representation of the pressure orientation. The attitude quaternion output by the attitude sensing unit of the upsetting head 3 is used to align the working surface coordinates to the gravity direction in the world coordinate system, so that the horizontal direction is fixed as the global reference. This alignment only changes the reference zero point of the angle and does not change the physical magnitude of the pressure, thus providing a unified coordinate reference for subsequent judgments under different installation postures.

[0030] After establishing the coordinate reference, the system registers sensor and channel information. The serial numbers of each sensor in the pressure sensor array 4 are marked as k=1, 2, ..., K. A mapping table from hardware to logical channels is established, and time synchronization detection is performed on all channels. The control cycle Ts=10ms and the timestamp strategy are established to ensure that multi-channel data enters the same time base ring buffer.

[0031] Subsequently, a no-load calibration was performed, placing the upsetting head 3 in a state without external force, and continuously collecting raw data for a period of time to determine the zero-point offset b of each channel. k With noise level σ k The normal and tangential channels are filed independently. The zero point is used for baseline removal during operation. The noise level is used to set the normalization weight and minimum contact weight threshold. If drift or abnormality occurs during no-load, the system determines that the channel is unqualified and prompts for maintenance.

[0032] Based on the no-load calibration, load calibration was performed to establish the conversion relationship between electrical readings and physical pressure. The normal channel was loaded stepwise with a known normal load, the steady-state response was measured, and the normal conversion factor C was determined. n，k The tangential channel is calibrated using a known tangential load, and the tangential conversion factor C is obtained. t，k Conversion factor and effective stress area A per piece k They are solidified and stored together, forming the resultant normal force F for subsequent calculations. n Tangential resultant force F t A metering chain for estimating the coefficient of friction.

[0033] System loading safety thresholds and preference parameters: On the angle side, the preferred pressure zone center and half-width are set, and the upsetting head working surface is divided radially into an inner region, a middle region, and an outer region, with the middle region set as the preferred pressure zone; on the pressure side, a safety upper limit F is set. max =100kN and minimum effective pressure threshold F min =1kN; Time-side configuration includes control period Ts and debouncing smoothing coefficient α. f =0.1, Resident statistics window N win =10, Shortest dwell time T min =0.5s and safe clearance time T clear =2s.

[0034] The system employs an online attitude and temperature compensation strategy. Attitude compensation continuously uses the output of the attitude sensing unit to align the working surface angle. Temperature compensation corrects the zero point based on the difference between the no-load reference temperature and the current temperature. The health status of the channel is periodically monitored, and if an abnormality is detected, it enters a conservative mode.

[0035] Finally, the system writes the calibration constant, threshold parameters, time configuration and alignment reference into non-volatile memory to form a data structure for runtime calls. When all channels pass the self-test, the system is set to an available state and proceeds to step S2.

[0036] Step S2: Data Acquisition and Pressure Location

[0037] The system synchronously acquires data from the pressure sensor array 4 over a control cycle Ts, including pressure estimation and attitude quaternions. Each sensor provides the normal pressure p. n，k With tangential shear p t，k Two channels, data timestamps are aligned and written to a circular buffer; when samples are lost, interpolation is used to fill in the gaps.

[0038] Based on the zero point and noise baseline obtained in step S1, the original readings are subjected to baseline removal and normalization. The normalization weights are calculated as follows:

[0039]

[0040] Total weight When W tot Below the minimum contact weight threshold W min When the value is 0.1, it is determined that there is no contact in this cycle, and no positioning value is output.

[0041] During the effective contact period, the system calculates the pressure distribution deviation P: ;where `k contact The average value of the triggered sensor sequence number, `k pref To optimize the average value of the pressure zone sensor serial numbers, for example, if the trigger sensor serial numbers are 4 and 5, then `k` contact =4.5; the preferred zone sequence numbers are 6, 7, and 8, then k pref =7.

[0042] Step S2 outputs the pressure distribution deviation P and the resultant force estimate F^ with timestamps, and continuously monitors the channel health status.

[0043] Step S3: Joint determination of pressure deviation and resultant force

[0044] This step takes the pressure distribution deviation P and the resultant force estimate F^ output from step S2 as inputs. The resultant force estimate F^ is the normal resultant force, and the calculation formula is as follows:

[0045]

[0046] When F^ is between F min and F max When the time interval is between, calculate the adjustment factor Q:

[0047]

[0048] Where α=0.6 and β=0.4 are weighting coefficients, the larger Q is, the less suitable the current pressure distribution is, and the hydraulic cylinder position needs to be adjusted.

[0049] At the same time, determine the pressure distribution area: the inner area corresponds to serial numbers 1-8, the outer area corresponds to 17-24, and the middle area corresponds to 9-16. If the pressure is in the inner area, hydraulic cylinder 2 needs to be moved backward; if it is in the outer area, it needs to be moved forward; if it is in the middle area, pressure fine adjustment is given priority.

[0050] Step S4: Hierarchical Residential Dependency and Step Size Quantization

[0051] This step takes the adjustment coefficient Q and the region determination result as inputs, and calls the shortest dwell time T. min and safe cleanup duration T clear The system updates the threshold count in each cycle, only updating it when Q continuously exceeds the threshold H=0.8 for a duration of T. min At that time, control commands are generated.

[0052] The adjustment direction is determined by the region: the inner region moves backward, the outer region moves forward, and the middle region is fine-tuned. The movement step size ΔL is calculated as follows:

[0053] First, calculate the over-threshold amplitude. and tightness index Nominal step size Where k0=1, k Q =0.45, k T =0.55, actual step size L max =3 is the maximum step size constraint.

[0054] After the control command is issued, the system enters a latching state to avoid repeated triggering. The adjustment strategy prioritizes the minimum step size and gradually increases it until Q drops below the threshold.

[0055] Step S5: Reset, Record, and Conservative Strategy

[0056] After successful adjustment, the system performs a reset: clears the over-threshold count, releases the latch, and at T... clear When the event ends, the event is closed, and key event data is recorded to non-volatile memory, including adjustment direction, step size, P and F^ before and after adjustment, Q peak values ​​and duration.

[0057] If the channel is abnormal or the proportion of failed chips exceeds the limit, enter conservative mode: increase the threshold, limit the decision complexity, and prompt maintenance. If the attitude or temperature compensation is not up to standard, reduce the prompt frequency or prohibit operation.

[0058] Example 2, the device structure of this example is as follows Figure 1-2As shown, frame 1 is a welded frame with dimensions of 2000mm×1000mm×500mm; hydraulic cylinder 2 has a stroke of 1000mm and a pressure range of 0-100kN; upsetting head 3 has a diameter of 300mm and is made of mold steel; pressure sensor array 4 contains 24 thin-film sensors that are evenly distributed; control unit 5 uses an industrial PLC, which integrates non-volatile memory and communication interface.

[0059] During operation, the control unit 5 executes the above control method to adjust the position of the hydraulic cylinder 2 in real time to ensure uniform pressure distribution. The structural design enhances rigidity and durability, making it suitable for high-strength forging scenarios.

[0060] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0061] In the several embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or units may be electrical, mechanical, or other forms.

[0062] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0063] In addition, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0064] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A horizontal hydraulic upsetting mechanism with large stroke and real-time pressure feedback, comprising a frame (1), a horizontally arranged hydraulic cylinder (2), and an upsetting head (3) connected to the hydraulic cylinder (2), characterized in that, The upsetting head (3) is provided with a pressure sensor array (4) on its working surface. A control unit (5) is provided above the frame (1). The control unit (5) is connected to the pressure sensor array (4) and the hydraulic cylinder (2) in communication. The control unit (5) is configured to perform the following steps: establish a unified geometric and coordinate reference; calibrate the pressure sensor array (4) under no-load and load conditions, and preset the threshold and the shortest dwell time. Pressure data from each pressure sensor array (4) is periodically collected. The pressure sensor array (4) is configured to simultaneously detect normal pressure and tangential shear force. Pressure distribution deviation and resultant force estimation are calculated based on the pressure data. An adjustment coefficient is calculated by weighted summation of the pressure distribution deviation and resultant force estimation, and the adjustment coefficient is compared with a preset threshold. When the adjustment coefficient continuously exceeds the preset threshold and reaches the predetermined minimum dwell time, a control command containing the adjustment direction and movement step size is generated, and the hydraulic cylinder (2) is controlled to move. The working surface of the upsetting head (3) is divided radially into an inner region, a middle region, and an outer region. The middle region is set as the preferred pressure zone. The adjustment direction is determined based on the area of ​​pressure distribution on the upsetting head (3): if the pressure zone is the inner region, the direction is indicated to move backward; if it is the outer region, the direction is indicated to move forward; if it is the middle region, the pressure fine-tuning is given priority. After the adjustment is completed, a reset procedure is executed, and the relevant data of this event is recorded to non-volatile memory.

2. The horizontal hydraulic upsetting mechanism according to claim 1, characterized in that, The upsetting head (3) is connected to the hydraulic cylinder (2) via a flange, and a reinforcing rib is provided at the connection.

3. The horizontal hydraulic upsetting mechanism according to claim 1, characterized in that, The upsetting head (3) is equipped with an attitude sensing unit, which is used to output the attitude information of the upsetting head (3).

4. The horizontal hydraulic upsetting mechanism according to claim 1, characterized in that, The calculation of pressure distribution deviation includes: obtaining the identification information of each triggered pressure sensor array (4) and calculating its average value; calculating the average value of the identification information of the pressure sensor array (4) corresponding to the preset preferred pressure zone; the pressure distribution deviation is the difference between the above two average values.

5. The horizontal hydraulic upsetting mechanism according to claim 1, characterized in that, The generation of control commands, which includes adjusting the direction and moving step size, specifically includes: continuously monitoring the adjustment coefficient, and only confirming the need to generate control commands when it continuously exceeds the preset threshold and reaches the preset minimum dwell time; the moving step size is a discrete integer value, the magnitude of which is determined by the threshold exceedance and the tightness index, and an upper limit constraint is applied to this value; and generating control commands based on the adjustment direction and moving step size.

6. The horizontal hydraulic upsetting mechanism according to claim 5, characterized in that, After the control command is issued, the system enters a latching state, during which no new commands are generated repeatedly until the operation execution result is received or the safe clearing time is reached.

7. The horizontal hydraulic upsetting mechanism according to claim 6, characterized in that, The over-threshold amplitude is the amplitude by which the adjustment coefficient exceeds the preset threshold, and the combined force estimation is normalized to obtain the tightness index.

8. The horizontal hydraulic upsetting mechanism according to claim 1, characterized in that, The control unit (5) is also configured to enter a conservative mode when the channel health status is abnormal, thereby increasing the threshold and limiting the decision complexity.

Images