An adaptive propulsion driver based on chemical exothermic reaction driving

Abstract

The application discloses a self-adaptive propelling driver based on chemical energy release reaction driving. The driver comprises a non-magnetic outer cylinder and a coaxial inner cylinder assembly, a plurality of groups of toe mechanisms are connected to the lower end of the outer cylinder through a one-way hinge, flexible web membranes are connected between adjacent toes, electric permanent magnets are arranged on the inner wall of the outer cylinder corresponding to the toes, a chemical energy release reaction chamber, a soft film, a spark plug and a mixed gas storage chamber are arranged in the inner cylinder. A controller executes a conventional propelling main mode and an abnormal escape bottom mode: in the conventional mode, the electric permanent magnets are switched according to the cycle control of leg folding swing, ground contact expansion, support propelling and drag reduction recovery, so that low-resistance high-efficiency propelling is realized; when it is detected that the conventional propelling is invalid due to soft mud subsidence, strong adhesion or negative pressure adsorption, the chemical energy release reaction is triggered to generate instantaneous high-pressure gas to drive the soft film to drum outward, and additional pulse power is output to realize transient speed escape. The application takes into account the propelling efficiency in ordinary water bodies and the detachment capacity in extreme working conditions, and significantly prolongs the endurance and improves the reliability.

Description

Technical Field

[0001] This invention relates to a propulsion actuator, specifically an adaptive propulsion actuator based on chemically released reaction. Background Technology

[0002] In existing technologies, underwater robots typically employ disc-shaped, spherical, or wheel-like leg structures for their feet. In ordinary water bodies, increasing the contact area is often necessary to improve propulsion, but this leads to a significant increase in fluid resistance during leg swinging and retrieval. When walking on soft mud, while increasing the contact area can reduce the ground pressure, it also makes the robot more susceptible to stronger adhesion and localized adsorption resistance when pulling up.

[0003] While bionics has developed webbed feet, most existing bionic webs are passive, flexible structures. They cannot actively control deployment or retraction based on water resistance, substrate bearing capacity, and the degree of sinking, making it difficult to balance propulsion efficiency in ordinary water bodies with the high-resistance detachment requirements of soft mud. Especially in amphibious terrains such as tidal flats, silt not only generates frictional resistance but also creates a localized vacuum adsorption when the foot is lifted. This adsorption force often exceeds the continuous output capacity of conventional motors, making reliable detachment difficult once the foot sinks deep into the mud. Summary of the Invention

[0004] This invention aims to solve common problems faced by existing underwater robots in the processes of navigation and propulsion in ordinary water bodies and in operations with high resistance on soft bottoms. Firstly, in ordinary water environments such as rivers, lakes, reservoirs, and nearshore shallow seas, the feet experience significant fluid resistance during leg swinging, retrieval, and phase transitions. If the propulsion area remains too large for an extended period, it can easily lead to decreased propulsion efficiency and increased energy consumption. Secondly, in estuaries, tidal flats, and soft mud environments, in addition to fluid resistance, there is further superimposed subsidence, adhesion, and localized negative pressure adsorption, resulting in difficulty in leg detachment or even burial failure during the upward pull-out phase. The basic objective of this invention is not to rely on chemical energy release for propulsion in all operating conditions, but rather to utilize electrically driven, electro-permanent magnet-controlled, duck-web-like feet to achieve low-resistance swinging, gradual ground contact, and support propulsion in most ordinary water bodies and general bottom conditions. Only when the amount of subsidence, the magnitude of propulsion or retrieval resistance, and the number of consecutive retrieval failures indicate that conventional propulsion can no longer achieve detachment will chemical energy release for transient speed change propulsion be triggered to provide detachment capability.

[0005] This invention provides an adaptive propulsion actuator based on chemically released reaction, comprising:

[0006] An outer cylinder body and an inner cylinder assembly coaxially sleeved inside it, with a sealed isolation structure between the outer cylinder body and the inner cylinder assembly. A dry zone receiving cavity is formed inside the outer cylinder body, and a reaction pressure-bearing space is formed inside the inner cylinder assembly.

[0007] Multiple toe mechanisms are evenly hinged to the lower outer periphery of the outer cylinder body via one-way hinges. A flexible web membrane connects adjacent toe mechanisms. A magnetic steel sheet is embedded on the back of each toe mechanism at the retracted position. A torsion spring that provides rebound and reset torque is installed at the hinge.

[0008] An electro-permanent magnet sheet is fixedly installed inside the outer cylinder wall at the position corresponding to each toe mechanism, and is connected to the webbed foot control circuit sealed inside the outer cylinder. It switches between magnetization and demagnetization states by applying a pulse current to attract or release the magnetic steel sheet.

[0009] A chemical energy release reaction chamber is disposed inside the inner cylinder assembly. The bottom drive end of the inner cylinder assembly is provided with a soft membrane, and the top is provided with a spark plug that extends into the chemical energy release reaction chamber.

[0010] A mixed gas storage chamber is located on the inner wall of the outer cylinder and is connected to the chemical energy release reaction chamber via an inlet pipe.

[0011] An exhaust pipe connects the chemical energy release reaction chamber to a one-way duckbill valve located above the chemical energy release reaction chamber, and the outlet of the one-way duckbill valve is connected to the exhaust channel at the top of the outer cylinder.

[0012] A chemical energy release reaction control circuit is sealed and installed inside the outer cylinder, and connected to the spark plug and the intake control terminal of the mixed gas storage chamber; and

[0013] The controller, connected to the webbed foot control circuit and the chemical energy release reaction control circuit, is configured to perform coordinated control of the conventional propulsion main mode and the abnormal escape and bottom-line mode: in the conventional propulsion main mode, the electro-permanent magnet plate is controlled to switch magnetization states according to the cyclic logic of leg retraction swing, ground contact gradual advance, support propulsion, and drag reduction recovery / leg withdrawal; when it is detected that conventional propulsion cannot achieve desorption, it switches to the abnormal escape and bottom-line mode, controls the chemical energy release reaction control circuit to quantitatively deliver the mixed medium to the chemical energy release reaction chamber and triggers the spark plug ignition, generating instantaneous high-pressure gas to drive the soft diaphragm to rapidly bulge outward to output additional pulse power.

[0014] Preferably, the toe mechanism consists of three pieces arranged side by side along the width of the web surface. The controller controls the synchronous or time-sequential deployment and retraction of the three toes according to the target deployment area and retraction resistance requirements.

[0015] Preferably, when using time-sharing deployment or time-sharing retraction, the control delay between adjacent toes is 30ms to 80ms to reduce the impact of a single complete deployment or retraction on the water body or soft bottom surface.

[0016] Preferably, in the conventional propulsion main mode, the controller cyclically controls multiple toe mechanisms in the sequence of leg retraction and swing, gradual ground contact and opening, support propulsion, and drag reduction retraction / leg withdrawal; wherein,

[0017] During the leg retraction and swing phase, the three toes are magnetically charged and retracted, so that the webbed feet are pressed tightly against the outer tube wall to reduce fluid resistance;

[0018] During the ground contact and gradual unfolding phase, the three toes are demagnetized and released synchronously or in stages according to the target unfolding area, so that the webbed feet unfold under the action of the torsion spring and form a support area that is compatible with the propulsion requirements and the bearing conditions of the bottom.

[0019] During the support and propulsion phase, the extended toes are kept in an extended state to output propulsive traction.

[0020] During the drag reduction recovery / leg removal stage, the corresponding electro-permanent magnet plate outputs a magnetizing pulse to make the three toes retract synchronously or in stages, thereby reducing the fluid resistance during the recovery process or the interface adhesion resistance under soft mud conditions.

[0021] The present invention also provides a control method for an adaptive propulsion actuator driven by the aforementioned chemical exothermic reaction, comprising the following steps:

[0022] Step S1, State Acquisition and Feature Estimation: Acquire foot end normal load, foot end axial displacement, body attitude, actual displacement increment, propulsion speed and reaction chamber pressure and filter them to estimate subsidence amount, slip ratio, propulsion / recovery resistance characterization quantity and high drag risk index;

[0023] Step S2, Working State Determination: Determine the contact state based on the dual threshold hysteresis criterion of normal load, and determine the working state of the actuator based on the characteristic quantity, including leg retraction swing, ground contact gradual development, support propulsion, drag reduction recovery / leg removal, interface release, chemical energy release transient speed change escape and exhaust reset state.

[0024] Step S3, conventional propulsion control: The toes are controlled to retract and extend according to the cyclic logic of leg retraction and swing, gradual extension upon ground contact, support propulsion, and drag reduction / leg withdrawal. During the leg retraction and swing phase, the toes are magnetized to reduce fluid resistance. During the gradual extension upon ground contact phase, the target extension area is determined based on the normal load, resistance characteristics, and settlement amount and mapped to the number of toes to be extended. During the support propulsion phase, the extension state is maintained and traction force is output. During the drag reduction / leg withdrawal phase, the toes are magnetized to reduce recovery resistance or interface adhesion resistance.

[0025] Step S4, Interface release control: In the drag reduction recovery / leg pulling state, when the recovery pull-up command is valid, the normal load is not higher than the release threshold, the settlement amount is greater than zero and does not exceed the shallow settlement threshold, and the settlement change rate is less than zero, a tangential reciprocating disturbance is applied to the toe tip to weaken the mud-water interface adhesion.

[0026] Step S5, Chemical Energy Release Instantaneous Speed ​​Escape Control: When conventional propulsion and interface deadhesion fail to detach, and the high resistance risk index, sinking amount, normal load or number of consecutive failures reach the bottom-line trigger threshold, and the body attitude and initial pressure of the reaction chamber meet the safety conditions, a mixed medium is quantitatively delivered to the chemical energy release reaction chamber and the spark plug is triggered to ignite, generating instantaneous high-pressure gas to drive the soft membrane outer drum to output additional pulse power.

[0027] Step S6, Exhaust Reset and Repeated Trigger Suppression: When the reaction chamber pressure is lower than the reset threshold, the pressure change rate is lower than the change rate threshold, and the normal load is lower than the release threshold, the state of exhaust reset is determined; and the time interval between two adjacent chemical energy release triggers is constrained to be no less than the minimum time interval.

[0028] Preferably, in step S2, the contact state adopts a dual-threshold hysteresis criterion: when the normal load is not lower than the upper threshold for grounding determination, it is determined to be grounding; when it is not higher than the lower threshold for leaving the ground determination, it is determined to be leaving the ground. The upper threshold is greater than the lower threshold to suppress frequent state switching caused by contact jitter.

[0029] Preferably, in step S3, the target unfolded area is determined by weighting the normal load, propulsion / recovery resistance and subsidence to determine the conventional target support area, and is obtained by introducing risk compensation and limiting the amplitude based on the deviation between the high resistance risk index and the risk threshold; the number of toe pieces to be unfolded is determined by mapping the target unfolded area to the proportion of the total number of toe pieces and rounding up, thereby realizing the conversion from continuous area requirements to discrete toe control.

[0030] Preferably, in step S4, the micro-vibration frequency, micro-vibration cycle number, and tangential disturbance amplitude of the tangential reciprocating disturbance are all adaptively determined based on the high resistance risk index and / or slip ratio, so that the release intensity is automatically adjusted according to the degree of adhesion risk.

[0031] Preferably, in step S6, the exhaust reset state must simultaneously meet three joint criteria: the reaction chamber pressure is lower than the reset threshold, the reaction chamber pressure change rate is lower than the change rate threshold, and the normal load is lower than the release threshold; and the time interval between two adjacent chemical energy release triggers is constrained by the minimum time interval to prevent continuous over-ignition.

[0032] Preferably, the method further includes step S7, safety interlocking and adaptive tuning: saving historical data from multiple recent control cycles, and adaptively tuning the release initiation threshold, abnormal fallback trigger threshold and minimum reference gas injection volume accordingly, so that the control parameters are automatically optimized as the operating conditions evolve.

[0033] The beneficial effects of this invention are:

[0034] This invention introduces electro-permanent magnet technology and chemical energy release instantaneous speed drive technology into the design of underwater robot feet, successfully solving the problems of high recovery resistance and limited propulsion efficiency in ordinary water bodies, as well as easy sinking, difficulty in leg removal, and insufficient instantaneous power in soft mud environments by introducing electro-permanent magnet technology and chemical energy release instantaneous speed drive technology.

[0035] The biomimetic duck webbed foot structure in this invention relies on electro-permanent magnet group control to complete low-resistance swinging, ground-touch deployment, support propulsion, and drag-reducing recovery / leg pulling under most working conditions, thereby achieving stable and low-energy conventional propulsion; when encountering extreme high-resistance working conditions such as soft mud burial, strong adhesion, or local negative pressure adsorption, chemical energy release drive can quickly destroy the high-resistance contact interface and output high-power-density pulse power to achieve a high success rate of abnormal escape.

[0036] By introducing a hierarchical adaptive algorithm, this invention can also distinguish between two working conditions: increased resistance in ordinary water bodies and sludge burial and desorption. This reduces the false triggering rate and gas consumption, and improves propulsion stability, continuous operation capability, and overall project reliability. Attached Figure Description

[0037] Figure 1 A schematic diagram of a multi-toed unidirectional unfolding bionic webbed foot propulsion component;

[0038] Figure 2 This is a cross-sectional schematic diagram of a chemical energy release reaction actuator;

[0039] Figure 3 This is the overall flowchart of the collaborative driving algorithm for propulsion in ordinary water bodies and desorption from soft, high-resistivity mud. Detailed Implementation

[0040] To make the objectives, technical solutions, and advantages of the present invention clearer, the present invention will be further described in detail below with reference to the accompanying drawings.

[0041] like Figure 1 and Figure 2 As shown, this embodiment provides an adaptive propulsion actuator based on chemical energy release reaction, the core of which adopts a double-cylinder coaxial sealing structure consisting of an outer cylinder body 7 and an inner cylinder assembly 10 coaxially sleeved inside it.

[0042] The outer cylinder 7 serves as the main load-bearing shell, and its cylinder wall 1 is made of non-magnetic thin-walled high-strength material (such as lightweight high-strength aluminum alloy or 316L stainless steel) to withstand external water pressure loads, while allowing internal magnetic fields to penetrate and drive external mechanisms.

[0043] The outer cylinder 7 forms a dry zone cavity, and the inner cylinder assembly 10 forms an independent reaction pressure-bearing space. A sealed isolation structure is formed between the two, realizing the spatial separation between the control circuit dry zone and the chemical energy release reaction pressure-bearing zone.

[0044] Multiple sets of toe mechanisms 6 are evenly distributed around the lower outer periphery of the outer cylinder 7. In this embodiment, three toe mechanisms 6 are preferably arranged side by side along the width of the web surface. Each toe mechanism 6 is connected to the outer cylinder 7 by a one-way hinge. Adjacent toe mechanisms 6 are connected by a flexible web membrane 4 to form a continuous biomimetic web surface structure. When the toe mechanisms 6 are deployed, the force-bearing area is increased to improve propulsion efficiency, and when the toe mechanisms 6 are retracted, fluid resistance is reduced.

[0045] A torsion spring 5 is installed at the hinge of the toe mechanism 6 to provide a rebound torque for the toes, so that the toes automatically return to their unfolded position after release. A magnetic steel sheet is embedded on the back of each toe mechanism 6 at the retracted position.

[0046] An electro-permanent magnet 3 is fixedly installed inside the outer cylinder 7 at the position corresponding to each toe mechanism 6. The electro-permanent magnet 3 is connected to the webbed foot control circuit in the webbed foot control circuit chamber 18 sealed inside the outer cylinder 7. The webbed foot control circuit switches between "magnetized" and "demagnetized" states by applying a pulse current to the electro-permanent magnet 3: In the magnetized state, the magnetic field penetrates the outer cylinder wall 1 and attracts the magnetic steel sheet on the blade, overcoming the restoring force of the torsion spring 5 and retracting the toes tightly against the outer cylinder wall 1, and the attraction state can be maintained without the need for current; in the demagnetized state, the magnetic force disappears, and the toes naturally spring open under the action of the torsion spring 5. The one-way hinge of the toe mechanism 6 provides a limiting function, ensuring that the toes can only be extended within a preset angle range.

[0047] In terms of chemical energy release drive, the inner cylinder assembly 10 has a chemical energy release reaction chamber 9 inside, which is the place where the chemical reaction occurs and high-pressure gas is generated. A large deformation is generated by a soft membrane 13 (which also serves as a flexible interface to isolate the reaction zone from the external environment) to achieve instantaneous speed drive, providing additional pulse power in high-resistance desorption conditions such as soft mud burial, strong adhesion, or other conventional recycling failures. The bottom drive end of the inner cylinder assembly 10 is equipped with a soft membrane 13, and the top is equipped with a spark plug 12 that extends into the chemical energy release reaction chamber 9. The inner wall of the outer cylinder body 7 is equipped with a mixed gas storage chamber 8, which is used to pre-store the mixed gas (such as propane and oxygen) required for the chemical energy release reaction. The mixed gas storage chamber 8 is connected to the chemical energy release reaction chamber 9 via an inlet pipe 15. A one-way duckbill valve 11 is connected to the top of the chemical energy release reaction chamber 9 via an outlet pipe 14. The outlet of the one-way duckbill valve 11 is connected to the outlet channel 16 at the top of the outer cylinder body 7, which is used to allow the waste gas after the reaction to be discharged in one direction and to prevent the backflow of external mud and water.

[0048] When the controller determines that the conventional propulsion cycle still cannot complete the recovery or desorption, and detects obvious entrapment / adhesion characteristics, the system switches to the abnormal escape and bottom-line mode. The mixed gas storage chamber 8 delivers a mixed medium to the chemical energy release reaction chamber 9 through the air intake pipe 15. After the spark plug 12 is ignited, it generates instantaneous high-pressure gas, which drives the soft membrane 13 to bulge outward rapidly, thereby providing additional pulse power to achieve high-resistance desorption.

[0049] The outer cylinder 7 has a pressure relief hole 2 at its upper or end position to release excess gas or liquid when internal pressure fluctuations or changes in ambient pressure occur. The top of the outer cylinder 7 has a chemical energy release reaction control circuit chamber 17 and a webbed foot control circuit chamber 18 arranged in layers. The two are physically isolated from each other and are used to house the circuit elements that control the ignition pulse of the spark plug 12, the intake logic of the mixed gas storage chamber 8, and the pulse circuit and control algorithm module that drive the magnetic state switching of the electro-permanent magnet 3, respectively.

[0050] In the conventional propulsion main mode, the webbed foot control circuit switches the magnetization state by sending pulse current to the electro-permanent magnet 3 sealed inside the cylinder wall:

[0051] During the leg swing phase, the three toes are magnetically charged and retracted, so that the webbed feet are in close contact with the outer cylinder to reduce fluid resistance and recovery resistance in ordinary water bodies;

[0052] During the landing phase, the three toes are demagnetized and released synchronously according to the target unfolding area, so that the webbed feet unfold accordingly under the action of the torsion spring and form a support area that is compatible with the current propulsion requirements and the bearing conditions of the bottom.

[0053] During the propulsion phase, the selected number of deployment groups is maintained to output propulsion traction in ordinary water bodies and reduce ground pressure under soft bottom contact conditions;

[0054] During the drag-reducing recovery / leg pull phase, the toes are brought together again to prioritize completing the next cycle through low-resistance recovery or drag-reducing upward pull.

[0055] To accommodate both ordinary water bodies and high-resistivity conditions in soft mud, the system prioritizes area and timing adjustments within the conventional propulsion main mode based on the propulsion / recovery resistance and subsidence volume. Only when significant adhesion, shallow subsidence, or burial characteristics are detected will the system sequentially enter the interface release and chemical energy release transient speed-driven process to provide additional pulse power to complete desorption after conventional recovery fails.

[0056] This embodiment adopts a coordinated control scheme of "conventional propulsion main mode + abnormal escape backup mode" under the same drive architecture. The algorithm runs in discrete period Tc, where k is the discrete time index, corresponding to the kth sampling / control cycle when the controller runs in period Tc; the (k) appended to the variable indicates the value of the variable in the kth control cycle, and the (k-1) appended to the variable indicates the value of the variable in the previous control cycle.

[0057] During the leg swing phase, the three toes are magnetically retracted to reduce fluid resistance in ordinary water and recovery resistance before lifting off soft surfaces. After contact with the ground, the normal load Fz(k) at the foot tip, the propulsion / recovery resistance characteristic D(k), the sinking amount h(k), and the slip ratio are used to determine the dynamics. Calculate the unfolded area of ​​the target The number of toe plates to be deployed, Nopen(k), is determined, and the three toe plates are demagnetized and deployed synchronously or in stages according to the target deployment area. This provides an effective propulsion area in ordinary water bodies, reduces the ground pressure in soft-bottom conditions, and outputs propulsion traction.

[0058] When flow velocity disturbances, increased recovery resistance, or decreased propulsion efficiency occur in ordinary water bodies, the controller prioritizes adjusting... The timing of each stage, Nopen(k) and M1-M4, is used to maintain normal propulsion without triggering chemical energy release transient speed drive.

[0059] Only when conventional propulsion mode fails to achieve recovery or desorption in high-resistivity conditions such as soft mud, sludge, or other conditions with obvious burial or adhesion characteristics, will the controller first determine whether the foot has entered the unloaded upward lifting zone in drag-reducing recovery / leg-pulling state; only when Fz(k) is not higher than the adhesion release threshold Fvib, the subsidence amount h(k) is not higher than the shallow subsidence threshold hvib, and the subsidence change rate is... Only when the value is less than 0 will the interface release state M5 be executed, which applies a small tangential reciprocating disturbance to the tip of the toe to weaken the adhesion of the mud-water interface.

[0060] If the adhesion fails to detach after release, and the risk index R(k), settlement amount h(k), normal load Fz(k), or number of consecutive failures is not met, the risk index R(k), settlement amount h(k), normal load Fz(k), or the ... If the fallback trigger threshold is reached, the system switches to the abnormal fallback escape state M6, and calculates the target gas injection volume according to the optimized algorithm. Intake duration tg(k) and pre-charge pressure The chemical energy release chamber 9 is driven to release high-pressure gas instantaneously, which pushes the soft film 13 to bulge outward rapidly and form an additional escape impulse.

[0061] After desorption is completed, the controller continues to execute the exhaust reset and repeated trigger suppression logic to reduce the reaction chamber pressure Pr(k) back to a safe range and write the threshold tuning result back to the next control cycle, thereby achieving unified closed-loop coordination between ordinary water propulsion and soft mud high-resistance desorption.

[0062] In a specific example, the control method of this application is as follows:

[0063] Step S1: Status Acquisition and Filtering

[0064] Within each control cycle Tc, the controller synchronously acquires the foot end normal load Fz(k), foot end axial displacement xl(k), body pitch angle θ(k), and body roll angle. Signals such as the actual displacement increment Deltas(k) of the machine body, the actual propulsion speed v(k) of the machine body, the pressure Pr(k) of the reaction chamber, and the feedback status of the actuator.

[0065] Since the original measurement signal is easily affected by underwater disturbances and transient impacts, a first-order exponential filter is first applied to the acquired signal to improve the stability of subsequent state estimation. The filtering process is expressed as follows:

[0066]

[0067] In the formula, yf(k) is the filtered output of the kth control cycle, y(k) is the original signal acquired in the kth control cycle, yf(k-1) is the filtered output of the previous control cycle, and α is the filtering coefficient, preferably 0.6-0.85; the larger α is, the higher the weight of the current sampled value, and the smaller α is, the smoother the filtering result.

[0068] For the foot-end normal load signal, unless otherwise specified, Fz(k) in the following text represents the normal load value after the above filtering process.

[0069] After obtaining a stable normal load signal, a dual-threshold hysteresis criterion is used to determine the contact state of the normal load: when Fz(k) ≥ Fcon, it is determined to be grounded; when Fz(k) ≤ Fcoff, it is determined to be off-ground. Here, Fcon is the upper threshold for grounding determination, and Fcoff is the lower threshold for off-ground determination, and Fcon is greater than Fcoff to form a hysteresis interval and suppress frequent state switching caused by contact jitter.

[0070] Step S2: Environmental Feature Estimation

[0071] After completing the basic state acquisition, the controller further constructs characteristic quantities such as subsidence amount, slip ratio, propulsion / recovery resistance, and high resistance risk index.

[0072] To characterize the degree of indentation of the foot relative to the free, ungrounded state, the subsidence amount in the k-th control cycle is defined as follows:

[0073] In the formula, h(k) is the estimated value of the subsidence in the kth control cycle, xl(k) is the current position of the foot along the drive axis, x0 is the reference position of the foot when it is in a free and ungrounded state; max[0,·] indicates that the subsidence is 0 when the foot has not actually subsided.

[0074] To measure the deficit in actual thrust relative to expected thrust, the slip ratio is defined as:

[0075]

[0076] In the formula, λ(k) is the slip ratio of the kth control cycle, Deltas(k) is the actual displacement increment of the body in the cycle, Lcmd is the command step size of the cycle, and ε is a very small positive number to prevent the denominator from being 0; the larger λ(k) is, the more obvious the loss of the actual propulsion relative to the expected propulsion.

[0077] After obtaining the slip ratio, in order to uniformly describe under-propulsion and velocity deviation as drag indicators that can be used for control decisions, the propulsion / recovery drag characterization quantity is constructed as follows:

[0078]

[0079] In the formula, D(k) is the propulsion / recovery drag characterization quantity, which is used to comprehensively reflect the degree of decrease in propulsion efficiency, increase in incoming flow disturbance and increase in recovery drag; vref(k) is the target propulsion speed, v(k) is the actual propulsion speed, and c1 and c2 are weighting coefficients used to balance the contributions of the slip ratio term and the velocity deviation term to the drag characterization quantity.

[0080] Furthermore, to unify drag, subsidence, load, attitude, and failure history under the same risk criterion, a high-drag risk index is defined as follows:

[0081]

[0082] In the formula, R(k) is the high drag risk index; hmax is the normalized benchmark for subsidence, Fmax is the normalized benchmark for normal load, θmax and ϕmax are the upper limits of the allowable deviation of pitch angle and roll angle, respectively, nfail(k) is the number of consecutive failed recovery attempts, Nref is the normalized benchmark for the number of failures, and w1 to w6 are the weight coefficients of each risk factor.

[0083] Therefore, in ordinary water body scenarios, h(k) is usually close to 0, and D(k) mainly reflects the increase in fluid resistance, inflow disturbance or decrease in propulsion efficiency; in soft mud bottom scenarios, h(k), Fz(k) and D(k) often increase simultaneously, thus jointly characterizing the risk of burial and adhesion.

[0084] Step S3: Determine Working Status

[0085] The controller determines the current operating stage of the actuator based on the contact state obtained in step S1 and the characteristic quantities obtained in step S2. The operating states include: leg retraction and swing state M1, gradual ground contact and deployment state M2, support and propulsion state M3, drag reduction and recovery / leg removal state M4, interface release state M5, chemical energy release and instantaneous speed change for escaping state M6, and exhaust and reset state M7. M1 to M4 constitute the conventional main propulsion cycle suitable for ordinary water bodies and general bottom sediment conditions, while M5 to M7 constitute the auxiliary and bottom-covering cycles for conditions involving soft mud adhesion or burial.

[0086] Step S4: Control of webbed feet deployment and retraction during normal propulsion.

[0087] In the normal propulsion mode, the controller cyclically controls multiple toe mechanisms 6 in the sequence of "leg retraction swing M1 - ground contact gradual advance M2 - support propulsion M3 - drag reduction retraction / leg withdrawal M4".

[0088] During the leg swing phase, the three toes are magnetized and retracted to make the webbed feet adhere tightly to the outer cylinder, thereby reducing fluid resistance in ordinary water and recovery resistance before the soft bottom is pulled up. After contact with the ground, the target deployment area and the number of toes to be deployed are calculated based on the foot end normal load Fz(k), propulsion / recovery resistance characterization quantity D(k), sinking amount h(k) and slip ratio λ(k).

[0089] To determine the required foundation support area for the conventional propulsion phase based on normal load, drag level, and settlement degree, the conventional target support area is first defined as follows:

[0090]

[0091] In the formula, Adrive(k) is the conventional target support area in the kth control cycle, Amin is the minimum effective deployment area, Anom is the maximum deployment area allowed by conventional propulsion, Fnom is the normalized reference for conventional load, and b1, b2 and b3 are area allocation coefficients used to characterize the influence of normal load, resistance characterization and settlement on the support area, respectively.

[0092] When the system detects a high-resistance precursor, risk compensation is introduced based on Adrive(k) to obtain the final target unfolded area:

[0093] When R(k) ≥ R1, ;

[0094] when hour, .

[0095] In the formula, Astar(k) is the final target unfolded area after taking into account the anomaly precursor compensation, R1 is the risk threshold for entering the anomaly precursor compensation, Amax is the maximum allowed unfolded area, and c1 is the risk compensation coefficient.

[0096] After obtaining the target unfolded area in a continuous sense, it is mapped to the discrete control quantity of the number of toe pieces. Therefore, the number of toe pieces to be unfolded is preferably determined as follows:

[0097]

[0098] In the formula, Nopen(k) is the number of toe pieces to be unfolded in the kth control cycle, and its value is 1 to 3; ceil[·] is used to ensure that when the unfolded area of ​​the target increases, the controller can achieve area approximation in the form of discrete changes in the number of toe pieces.

[0099] In terms of control implementation, when the actuator is in the ground-contact gradual deployment state M2, the controller controls the three toes to demagnetize and deploy synchronously or in a time-sharing manner according to Astar(k) and Nopen(k). When using time-sharing deployment, a control delay of 30-80ms is set between adjacent toes to reduce the impact of a single full deployment on water disturbance or soft bottom surfaces and to suppress additional subsidence. When the actuator is in the support propulsion state M3, the controller keeps the deployed toes in the deployed state and maintains the propulsion posture until Deltas(k) ≥ Sref or the support time reaches Tsup; where Sref is the target displacement threshold for a single support propulsion stage, and Tsup is the maximum duration of the support propulsion stage. When the actuator is in the drag reduction recovery / leg withdrawal state M4, the controller outputs a magnetizing pulse to the corresponding electro-permanent magnet 3 according to the recovery or withdrawal requirements, causing the three toes to retract synchronously or in a time-sharing manner, thereby reducing the fluid resistance during the recovery process in ordinary water bodies and reducing the interface adhesion resistance during the withdrawal process in soft mud conditions.

[0100] Step S5: Interface release control

[0101] Interface deadsorption control is an auxiliary desorption step in the conventional propulsion main mode, and is mainly applicable to soft mud, sparse and soft sediments or other high-resistivity conditions with obvious interface adhesion.

[0102] In order to enable the intensity of the release action to adaptively adjust according to the degree of high resistance risk, the micro-vibration frequency is defined as follows:

[0103]

[0104] In the formula, fv(k) is the frequency of the release micro-vibration in the k-th control cycle, fmin and fmax are the minimum and maximum frequencies of the release micro-vibration, respectively, R1 is the low-risk threshold for release initiation, and R2 is the high-risk threshold before triggering the abnormal bottom-line escape. This represents a saturation function used to limit the input quantity to a preset normalization range.

[0105] After determining the vibration frequency, the number of micro-vibration cycles is defined as:

[0106]

[0107] In the formula, Nv(k) is the number of micro-vibration cycles applied in the k-th control cycle, N0 is the number of basic vibration cycles, and kn is the risk amplification factor, which is used to moderately increase the duration of micro-vibration under high-risk conditions as R(k) increases. This represents the function for rounding up.

[0108] Meanwhile, to enable the amplitude of the tangential disturbance at the toe tip to adapt to the slip state, it is preferable to express the amplitude of the tangential disturbance as follows:

[0109]

[0110] In the formula, deltatip(k) is the tangential disturbance amplitude of the toe tip in the kth control cycle, delta0 is the basic disturbance amplitude, kd is the slip ratio amplification factor, and deltamin and deltamax are the minimum and maximum values ​​of the disturbance amplitude, respectively. This represents the limiting function, used to constrain x within [...]. Within the range.

[0111] In satisfying Fz(k)≤Fvib and Only when the controller enters the interface release control. Here, ulift(k) is the recovery / lifting command of the current control cycle, Fvib is the load threshold that allows the release micro-vibration, and hvib is the upper limit of the shallow burial state; these conditions are used together to ensure that the release action occurs during the load reduction and lifting stage, rather than during the high load deep burial stage.

[0112] Step S6: Chemical energy release and transient speed change escape control

[0113] This step is triggered as an abnormal fallback strategy only when the support propulsion, drag reduction recovery / leg removal and interface release measures in the conventional propulsion main mode fail to complete the desorption. It is applicable to high-resistivity working conditions with obvious burial, adhesion or local negative pressure adsorption characteristics.

[0114] To adaptively determine the injection intensity based on the degree of burial, resistance level, foot load, and external hydrostatic pressure, the target mixed gas injection quantity is calculated as follows:

[0115]

[0116] In the formula, Vgstar(k) is the target mixed gas injection amount for the kth control cycle, Vmin and Vmax are the minimum and maximum allowable gas injection amounts, respectively, a1 to a4 are the gas injection amount allocation coefficients, Pw(k) is the current external hydrostatic pressure, and Pwref is the hydrostatic pressure normalization benchmark.

[0117] Once the target injection volume is determined, in order to convert the volume control quantity into a time control quantity that the actuator can directly call, the corresponding intake duration can be expressed as:

[0118]

[0119] In the formula, tg(k) is the intake time corresponding to the target injection volume, and qg is the calibrated flow rate of the intake pipe 15; after the intake pipe structure and supply pressure are determined, qg can be obtained through experimental calibration.

[0120] To avoid exceeding the peak pressure limit in the reaction chamber after ignition, the reaction chamber pressure is predicted before gas injection is completed and ignition is implemented:

[0121]

[0122] In the formula, Ppre(k) is the predicted pressure for the kth control cycle, P0 is the initial pressure of the reaction chamber, Kg is the conversion factor from injection volume to pressure rise, and Vr is the equivalent volume of the reaction chamber. This predictive relationship is used to assess whether the peak pressure may exceed the limit before ignition. When Ppre(k) > Plim, according to... The gas injection volume is limited, where Plim is the upper limit of the allowable reaction chamber pressure.

[0123] Regarding the triggering condition, it is preferable that R(k) ≥ R2 only. or , nfail(k)≥Nfail, and simultaneously satisfy the body attitude |θ(k)|≤thetae, Only when the reaction chamber pressure Pr(k) ≤ Parm and the pre-ignition interlock check is passed, is it permissible to switch to the abnormal bottom-out escape mode. Here, Fdeep is the load criterion for the deep-sinking state, Nfail is the threshold for the number of consecutive failures, thetae and phie are the upper limits of the allowable pitch and roll angles before triggering ignition, respectively, and Parm is the upper limit of the allowable initial reaction chamber pressure before ignition.

[0124] Step S7: Exhaust Reset and Repeated Trigger Suppression

[0125] After completing a chemical energy release transient speed-up drive, the controller jointly determines the reaction chamber pressure, the exhaust status of the outlet channel, and the foot release status to confirm whether the system has returned to a safe state to enter the next cycle.

[0126] When Pr(k) ≤ Preset, When Fz(k)≤Frel, it can be determined that the reaction chamber has completed effective depressurization and entered the exhaust reset state M7.

[0127] In the formula, Preset is the allowable pressure threshold during the exhaust reset phase. γp represents the rate of change of the reaction chamber pressure within adjacent control cycles, γp is the pressure change rate threshold, and Frel is the release load threshold after foot desorption.

[0128] To prevent continuous over-ignition during a single escape cycle, a minimum time interval is set between two adjacent chemical energy release triggers, and the constraint relationship is expressed as follows:

[0129]

[0130] In the formula, tfire(i) and tfire(i+1) represent the trigger times of the i-th and i+1-th chemical energy release, respectively, and Tlock is the minimum time interval between two adjacent triggers. By setting Tlock, structural overload or energy waste caused by continuous over-ignition can be avoided.

[0131] Step S8: Safety Interlocking and Adaptive Tuning

[0132] The controller stores the amount of sinking, propulsion / recovery resistance, chemical escape success rate, peak pressure in the reaction chamber, and mixed gas consumption of the most recent N steps, and adaptively tunes the viscosity release threshold, chemical escape trigger threshold, and baseline gas injection volume accordingly.

[0133] To enable the release threshold to be progressively adjusted based on recently successfully released samples, the following update relationship is adopted:

[0134]

[0135] In the formula, R1new and R1old are the updated release initiation risk threshold and the original release initiation risk threshold, respectively, β is the threshold update coefficient, and Rbarrelease is the average risk index corresponding to the recent successful release / desorption samples.

[0136] Accordingly, to enable the fallback trigger threshold to be adaptively adjusted based on recent failure samples, the following update relationship is adopted:

[0137]

[0138] In the formula, R2new and R2old are the updated and unupdated risk thresholds for triggering anomalies, respectively, and Rbarfail is the average risk index corresponding to recent samples of release failure or high resistance failure.

[0139] In addition, to enable the baseline gas injection volume to adaptively adjust with changes in recent chemical escape success rates, the minimum baseline gas injection volume is updated as follows:

[0140]

[0141] In the formula, Vminnew and Vminold are the updated minimum baseline gas injection volume and the original minimum baseline gas injection volume, respectively; μ is the gas injection volume adjustment coefficient; ηrelease is the recent chemical escape success rate; and Vlow and Vhigh are the lower and upper limits of the allowable variation range of the minimum baseline gas injection volume, respectively.

[0142] Based on the above steps Figure 3 It can be further refined into an integrated closed-loop process of "state acquisition and filtering - environmental characteristic quantity estimation - working state determination - M1 leg retraction and swing - M2 ground contact and gradual development - M3 support propulsion - M4 drag reduction and recovery / leg removal - high resistance source identification - M5 interface release - M6 abnormal bottoming and escape - M7 exhaust reset and parameter self-adaptation - return to the next control cycle", achieving unified closed-loop coordination between ordinary water body propulsion and soft mud high resistance desorption.

[0143] This embodiment is applicable to ordinary aquatic environments such as rivers, lakes, reservoirs, and nearshore shallow seas, as well as transitional areas of estuaries, tidal flats, and soft mud bottoms. The outer cylinder structure of the actuator is preferably made of lightweight, high-strength aluminum alloy or 316L stainless steel to adapt to environments with high oxygen content, high humidity, and salt corrosion. In ordinary aquatic scenarios, the actuator mainly faces the challenge of balancing the resistance of leg recovery, inflow disturbance, and propulsion efficiency; in soft mud bottom scenarios, the high-resistance desorption problem caused by settling, adhesion, and local negative pressure adsorption is further compounded.

[0144] Under most operating conditions, the control system operates in the conventional propulsion mode, using electro-permanent magnet group control to achieve low-resistance oscillation, gradual ground contact deployment, supported propulsion, and drag-reducing recovery / leg removal. When encountering extreme high-resistance conditions such as burrowing in soft mud, strong adhesion, or localized negative pressure adsorption, chemical energy release can rapidly disrupt the high-resistance contact interface and output high-power-density pulsed power, achieving a high success rate for escaping from abnormal situations. By introducing the aforementioned layered adaptive algorithm, the system can distinguish between two types of operating conditions: increased resistance in ordinary water bodies and detachment from soft mud burrowing, reducing false triggering rates and fuel consumption, and improving propulsion stability, continuous operation capability, and overall engineering reliability.

[0145] In summary, this invention utilizes electro-permanent magnet technology, consuming electrical energy only during the brief transition between toe retraction and extension states, significantly reducing system energy consumption and extending the robot's endurance. Since all core electrical components are sealed within a pressure-resistant outer cylinder, and the connection circuit with the electro-permanent magnet is also built into the toe, the reliability of the device in sandy waters is greatly improved. Furthermore, this invention integrates conventional foot propulsion, drag reduction / recovery / leg removal, interface deadhesion, and chemical energy release for instantaneous speed-up escape in layers. This allows the device to achieve efficient propulsion and recovery in ordinary water environments with low energy consumption, and in high-resistance conditions such as soft mud and loose sediments, the chemical energy release module is triggered only briefly when necessary, thus balancing propulsion efficiency, deadhesion capability, and energy utilization.

[0146] This invention, by integrating chemical energy release drive and electro-permanent magnet magnetic control technology, demonstrates significant technical advantages in the fields of propulsion in ordinary water bodies and high-resistance operations in soft mud. Firstly, in ordinary water environments, by actively controlling the deployment and retraction of the webbed feet, this invention can provide an effective propulsion area during the support phase and significantly reduce fluid resistance during the leg swing and recovery phases, thereby improving gait efficiency and endurance. Secondly, in extreme soft-bottom environments such as deep silt in tidal flats, the explosive power generated by the chemical energy release reaction chamber, coupled with the large deformation of the soft membrane, can instantly disrupt the mud-foot interface contact, offset adsorption resistance, and provide high initial acceleration when conventional electric drive cannot achieve desorption, enabling instantaneous speed escape.

[0147] In terms of energy utilization and system reliability, this invention utilizes electro-permanent magnet technology to maintain low power consumption during normal propulsion, consuming energy only during the moment of toe retraction and extension, significantly extending the robot's endurance. Since all core electrical components and the ignition control system are sealed within the pressure-resistant outer cylinder dry zone, and the external toe mechanism is controlled by electro-permanent magnets, this device completely eliminates the risk of leakage due to erosion from mud and sand inherent in traditional dynamic sealing structures, greatly improving reliability in sandy and corrosive waters. Furthermore, by organizing the normal propulsion, interface release, and chemical energy release instantaneous speed-change escape functions into primary and backup modes, this invention achieves a balance between daily propulsion efficiency and extreme sinking and escape capabilities without continuously maintaining high power output under all operating conditions.

[0148] Finally, this invention possesses exceptional environmental adaptability and continuous operation capability. Its biomimetic duck-web structure, through a composite control of active electronic control and passive fluid adaptation, enables low-resistance oscillation, stable propulsion, and efficient recovery in ordinary water bodies. Furthermore, in high-resistance substrates such as soft mud and dilute sediments, it gradually reduces ground pressure and achieves abnormal desorption through interfacial deadsorption and transient speed-driven processes. Combined with the conventional propulsion main cycle, high-resistance source identification, interfacial deadsorption control, abnormal escape triggering, quantitative gas injection ignition, and exhaust reset strategies in the stratified adaptive algorithm, this device can achieve stable operation under varying flow velocities, water contents, viscosities, and hydrostatic pressures.

[0149] The embodiments described above are merely preferred embodiments of the present invention and are not intended to limit the scope of the present invention. Various modifications and improvements made to the technical solutions of the present invention by those skilled in the art without departing from the spirit of the present invention should fall within the protection scope defined by the claims of the present invention.

Claims

1. An adaptive propulsion actuator based on chemically released reaction, characterized in that, include: An outer cylinder body and an inner cylinder assembly coaxially sleeved inside it, with a sealed isolation structure between the outer cylinder body and the inner cylinder assembly. A dry zone receiving cavity is formed inside the outer cylinder body, and a reaction pressure-bearing space is formed inside the inner cylinder assembly. Multiple toe mechanisms are evenly hinged to the lower outer periphery of the outer cylinder body via one-way hinges. A flexible web membrane connects adjacent toe mechanisms. A magnetic steel sheet is embedded on the back of each toe mechanism at the retracted position. A torsion spring that provides rebound and reset torque is installed at the hinge. An electro-permanent magnet sheet is fixedly installed inside the outer cylinder wall at the position corresponding to each toe mechanism, and is connected to the webbed foot control circuit sealed inside the outer cylinder. It switches between magnetization and demagnetization states by applying a pulse current to attract or release the magnetic steel sheet. A chemical energy release reaction chamber is disposed inside the inner cylinder assembly. The bottom drive end of the inner cylinder assembly is provided with a soft membrane, and the top is provided with a spark plug that extends into the chemical energy release reaction chamber. A mixed gas storage chamber is located on the inner wall of the outer cylinder and is connected to the chemical energy release reaction chamber via an inlet pipe. An exhaust pipe connects the chemical energy release reaction chamber to a one-way duckbill valve located above the chemical energy release reaction chamber, and the outlet of the one-way duckbill valve is connected to the exhaust channel at the top of the outer cylinder. A chemical energy release reaction control circuit is sealed and installed inside the outer cylinder, and connected to the inlet control terminal of the spark plug and the mixed gas storage chamber; as well as The controller, connected to the webbed foot control circuit and the chemical energy release reaction control circuit, is configured to perform coordinated control of the conventional propulsion main mode and the abnormal escape and bottom-line mode: in the conventional propulsion main mode, the electro-permanent magnet plate is controlled to switch magnetization states according to the cyclic logic of leg retraction swing, ground contact gradual advance, support propulsion, and drag reduction recovery / leg withdrawal; when it is detected that conventional propulsion cannot achieve desorption, it switches to the abnormal escape and bottom-line mode, controls the chemical energy release reaction control circuit to quantitatively deliver the mixed medium to the chemical energy release reaction chamber and triggers the spark plug ignition, generating instantaneous high-pressure gas to drive the soft diaphragm to rapidly bulge outward to output additional pulse power.

2. The adaptive propulsion actuator based on chemically released reaction as described in claim 1, characterized in that, The toe mechanism consists of three pieces arranged side by side along the width of the webbed surface. The controller controls the synchronous or time-sequential deployment and retraction of the three toes according to the target deployment area and retraction resistance requirements.

3. The adaptive propulsion actuator based on chemically released reaction as described in claim 2, characterized in that, When using time-sharing deployment or time-sharing retraction, the control delay between adjacent toes is 30ms to 80ms to reduce the impact on water disturbance or soft bottom surfaces caused by a one-time full deployment or retraction.

4. The adaptive propulsion actuator based on chemically released reaction as described in claim 1, characterized in that, In the conventional propulsion mode, the controller cyclically controls multiple toe mechanisms in the sequence of leg retraction and swing, gradual ground contact and opening, support propulsion, and drag reduction retraction / leg withdrawal; wherein, During the leg retraction and swing phase, the three toes are magnetically charged and retracted, so that the webbed feet are pressed tightly against the outer tube wall to reduce fluid resistance; During the ground contact and gradual unfolding phase, the three toes are demagnetized and released synchronously or in stages according to the target unfolding area, so that the webbed feet unfold under the action of the torsion spring and form a support area that is compatible with the propulsion requirements and the bearing conditions of the bottom. During the support and propulsion phase, the extended toes are kept in an extended state to output propulsive traction. During the drag reduction recovery / leg removal stage, the corresponding electro-permanent magnet plate outputs a magnetizing pulse to make the three toes retract synchronously or in stages, thereby reducing the fluid resistance during the recovery process or the interface adhesion resistance under soft mud conditions.

5. A control method for an adaptive propulsion actuator driven by a chemically exothermic reaction as described in any one of claims 1 to 4, characterized in that, Step S1, State Acquisition and Feature Estimation: Acquire foot end normal load, foot end axial displacement, body attitude, actual displacement increment, propulsion speed and reaction chamber pressure and filter them to estimate subsidence amount, slip ratio, propulsion / recovery resistance characterization quantity and high drag risk index; Step S2, Working State Determination: Determine the contact state based on the dual threshold hysteresis criterion of normal load, and determine the working state of the actuator based on the characteristic quantity, including leg retraction swing, ground contact gradual development, support propulsion, drag reduction recovery / leg removal, interface release, chemical energy release transient speed change escape and exhaust reset state. Step S3, conventional propulsion control: The toes are controlled to retract and extend according to the cyclic logic of leg retraction and swing, gradual extension upon ground contact, support propulsion, and drag reduction / leg withdrawal. During the leg retraction and swing phase, the toes are magnetized to reduce fluid resistance. During the gradual extension upon ground contact phase, the target extension area is determined based on the normal load, resistance characteristics, and settlement amount and mapped to the number of toes to be extended. During the support propulsion phase, the extension state is maintained and traction force is output. During the drag reduction / leg withdrawal phase, the toes are magnetized to reduce recovery resistance or interface adhesion resistance. Step S4, Interface release control: In the drag reduction recovery / leg pulling state, when the recovery pull-up command is valid, the normal load is not higher than the release threshold, the settlement amount is greater than zero and does not exceed the shallow settlement threshold, and the settlement change rate is less than zero, a tangential reciprocating disturbance is applied to the toe tip to weaken the mud-water interface adhesion. Step S5, Chemical Energy Release Instantaneous Speed ​​Escape Control: When conventional propulsion and interface deadhesion fail to detach, and the high resistance risk index, sinking amount, normal load or number of consecutive failures reach the bottom-line trigger threshold, and the body attitude and initial pressure of the reaction chamber meet the safety conditions, a mixed medium is quantitatively delivered to the chemical energy release reaction chamber and the spark plug is triggered to ignite, generating instantaneous high-pressure gas to drive the soft membrane outer drum to output additional pulse power. Step S6, Exhaust Reset and Repeated Trigger Suppression: When the reaction chamber pressure is lower than the reset threshold, the pressure change rate is lower than the change rate threshold, and the normal load is lower than the release threshold, the state of exhaust reset is determined; and the time interval between two adjacent chemical energy release triggers is constrained to be no less than the minimum time interval.

6. The control method according to claim 5, characterized in that, In step S2, the contact state adopts a dual-threshold hysteresis criterion: when the normal load is not lower than the upper threshold for grounding determination, it is determined to be grounding; when it is not higher than the lower threshold for leaving the ground determination, it is determined to be leaving the ground. The upper threshold is greater than the lower threshold to suppress frequent state switching caused by contact jitter.

7. The control method according to claim 6, characterized in that, In step S3, the target unfolding area is determined by weighting the normal load, the propulsion / recovery resistance, and the subsidence to determine the conventional target support area, and then the amplitude is obtained by introducing risk compensation based on the deviation between the high resistance risk index and the risk threshold. The number of toe pieces to be unfolded is determined by mapping the target unfolded area to the proportion of the total number of toe pieces and rounding up, thus realizing the transformation from continuous area requirements to discrete toe control.

8. The control method according to claim 6 or 7, characterized in that, In step S4, the micro-vibration frequency, micro-vibration cycle number, and tangential perturbation amplitude of the tangential reciprocating disturbance are all adaptively determined based on the high resistance risk index and / or slip ratio, so that the release intensity is automatically adjusted according to the degree of adhesion risk.

9. The control method according to claim 1, characterized in that, In step S6, the exhaust reset state must simultaneously meet three joint criteria: the reaction chamber pressure is lower than the reset threshold, the reaction chamber pressure change rate is lower than the change rate threshold, and the normal load is lower than the release threshold; and the time interval between two adjacent chemical energy release triggers is constrained by the minimum time interval to prevent continuous over-ignition.

10. The control method according to claim 1, characterized in that, It also includes step S7, safety interlocking and adaptive tuning: saving historical data from the most recent control cycles, and adaptively tuning the release initiation threshold, abnormal fallback trigger threshold and minimum reference gas injection volume accordingly, so that the control parameters are automatically optimized as the operating conditions evolve.

Images