A miniature servo motor cylinder

Abstract

The present application belongs to the field of precision driving and control technology, and relates to a micro servo electric cylinder, which comprises a fault judgment module, a push rod arranged from top to bottom, a hollow cup motor and a sensor flange, and the output shaft of the hollow cup motor and the push rod are parallel to the up-down direction; the sensor flange comprises a horizontally arranged disc body; four first strain gauges are arranged on the upper surface of the disc body, and the four first strain gauges are connected to form a bridge circuit I; four second strain gauges are arranged on the lower surface of the disc body, and the four second strain gauges are connected to form a bridge circuit II; the bridge circuit I and the bridge circuit II are both Wheatstone bridges connected in full-bridge mode; the first strain gauges and the second strain gauges are both used for reflecting the push-pull force received by the push rod; and the fault judgment module is used for detecting the faults of the bridge circuit I and the bridge circuit II. The present application solves the problem that the single face strain gauge fault leads to the failure of force control mode under the configuration of a single force sensor in the prior art.

Description

Technical Field

[0001] This invention belongs to the field of precision drive and control technology, and relates to a miniature servo electric cylinder. Background Technology

[0002] Miniature electric cylinders, as high-precision linear motion actuators, are widely used in industrial robots, medical equipment, and precision optical instruments. Their core technology lies in achieving high-precision force control within a small space through compact design. However, existing miniature electric cylinders still have significant shortcomings in force control reliability, limiting their applicability in high-end applications.

[0003] Force sensors are key components in the closed-loop control of electric cylinders, and their accuracy directly affects the force control performance of the actuator. Existing technologies often rely on single force sensor configurations. For example, the electric cylinder solutions with built-in force sensors disclosed in patents CN209329874U and CN222484494U, while achieving integrated force control, still employ a single sensor design. This sensor is placed directly inside the cylinder, occupying a large space and unsuitable for micro-cylinder applications. Furthermore, a malfunction of this single force sensor during use can cause the entire force control mode of the electric cylinder to fail. For instance, in high-precision applications such as humanoid robot dexterous hands, high-precision medical equipment, and microelectronic component press-fitting, a malfunctioning force sensor can lead to anything from interrupted force control to extreme force output values. This can result in dexterous hand grip failure, excessive gripping force causing crushing of objects, nerve damage in patients due to exceeding safe force values ​​in medical devices, and installation stagnation and damage to brittle materials due to force feedback mechanism failure in microelectronic press-fitting processes, causing production losses, equipment damage, and even safety accidents.

[0004] In summary, existing miniature electric cylinders rely on a single force sensor configuration. Once the sensor fails, the force control mode will fail, which in turn will cause the force control function to be unable to operate continuously and stably. There is an urgent need for a new technical solution to solve these problems. Summary of the Invention

[0005] The purpose of this invention is to solve the problems existing in the prior art and to provide a miniature servo electric cylinder.

[0006] To achieve the above objectives, the technical solution adopted by the present invention is as follows:

[0007] A miniature servo electric cylinder includes a fault diagnosis module and a push rod, a hollow cup motor, and a sensor flange arranged from top to bottom. The output shaft of the hollow cup motor and the push rod are both parallel to the vertical direction.

[0008] The sensor flange includes a horizontally arranged disc.

[0009] Four first strain gauges are provided on the upper surface of the disk, and the four first strain gauges are connected through resistors and amplifiers to form bridge circuit I; four second strain gauges are provided on the lower surface of the disk, and the four second strain gauges are connected through resistors and amplifiers to form bridge circuit II; bridge circuit I and bridge circuit II are both Wheatstone bridges with full bridge connection; the first strain gauges and the second strain gauges are used to reflect the push and pull forces on the push rod;

[0010] The fault diagnosis module is used to detect faults in Bridge Circuit I and Bridge Circuit II.

[0011] The force control accuracy of miniature servo electric cylinders is easily affected by temperature. The core problem lies in the conflict between the temperature drift error of the force sensor (such as strain gauge) signal chain and the physical limitations of miniaturization. The strain gauge generates a small current signal through mechanical deformation, which is converted into a voltage by a sampling resistor and then amplified by an amplifier. However, temperature changes can disrupt this process: First, the resistance of the sampling resistor increases with temperature, causing the output voltage to deviate from zero in the initial state (e.g., when the temperature rises by 50°C, the resistance of a metal film resistor increases by 0.5%, corresponding to a voltage shift of tens of millivolts); second, the bias current of the amplifier input stage increases with temperature, leading to distortion of the amplification ratio (e.g., when the temperature rises by 50°C, the bias current of a JFET amplifier increases by 50nA, resulting in a 0.1% error in the amplification factor).

[0012] In this invention, bridge circuit I and bridge circuit II constitute a redundant backup system. The two sets of bridge circuits serve as backups for each other, and a fault detection module detects whether either of them has failed. When one bridge circuit fails, the system can switch to the other bridge circuit for force measurement, thereby maintaining the operation of the micro servo electric cylinder's force control function and significantly improving the system's reliability. When both bridge circuits fail, a shutdown can be set to improve operational accuracy and prevent irreversible results caused by fault-induced malfunctions leading to incorrect force values ​​applied to the target object.

[0013] In existing technologies, patent CN219287324U does not contain a force sensor, thus failing to provide force control, and its internal FPC flexible cabling results in a large size. Patent applications CN117118131A and CN115987010A only use rotary encoders, lacking force sensing functionality and limiting their applications. Most existing technologies either lack a force sensor or only use full-bridge strain gauges on one side, resulting in a small strain range, low accuracy, and failure of force control due to the failure of a single full-bridge strain gauge. This invention solves the accuracy problem of single-sided full-bridge strain gauges by setting four strain gauges to form a Wheatstone bridge on the upper and lower surfaces of the disc, and avoids force control failure caused by single-point failure through redundant design.

[0014] As a preferred technical solution:

[0015] As described above, the fault diagnosis module of a miniature servo electric cylinder includes a first fault diagnosis module, which is used to detect open circuit faults in bridge circuit I and bridge circuit II. The detection principle is as follows: when an open circuit occurs, the bridge circuit output voltage will suddenly drop to 0, and at this time, the voltage residual value of the bridge circuit will increase and exceed the set threshold. The specific workflow is as follows:

[0016] (1) Initialize parameters: Let t=0 (initial time), set the sampling time interval Δt (Δt=0.01s-0.1s, the specific value can be customized by the user, and 0.05s is preferred in the embodiment);

[0017] (2) Collect the instantaneous voltage value of bridge circuit I at time t. Instantaneous voltage value output by bridge circuit II ;

[0018] (3) Calculate the predicted instantaneous voltage value at time t The residual of bridge I at time t The residual of bridge section II at time t The formula is as follows:

[0019] ;

[0020] ;

[0021] ;

[0022] In the formula, These are the preset fusion weight coefficients;

[0023] (4) Determine whether conditions ① and ② are both true. Condition ① is: and Condition ② is: and ;

[0024] If both conditions are met, then it is determined that both Bridge Road I and Bridge Road II are disconnected, and the program terminates.

[0025] If only condition ① is met, then it is determined that bridge road I is disconnected, and the measurement work is switched to bridge road II, and the program terminates;

[0026] If only condition ② is met, then it is determined that bridge road II is disconnected, and the measurement work is switched to bridge road I, and the program terminates;

[0027] If none of them are true, proceed to step (5);

[0028] (5) Update sampling time: Let t = t + Δt, and return to step (2).

[0029] As described above, the fault diagnosis module of a miniature servo electric cylinder includes a second fault diagnosis module, which is used to detect unidirectional drift faults in bridge circuit I and bridge circuit II. The detection principle is as follows: when unidirectional drift occurs in the circuit, the resistance of a single bridge arm will increase nonlinearly, which will also result in a unidirectional increase in the bridge voltage residual. To avoid the uncertainty caused by a single sudden change, this invention uses the method of calculating the moving average of the voltage residual to improve the accuracy of the judgment. The specific workflow is as follows:

[0030] (1) Initialize parameters: Let t=0 (initial time), set the sampling time interval Δt (Δt=0.01s-0.1s, the specific value can be customized by the user, preferably 0.05s);

[0031] (2) Collect the instantaneous voltage value of bridge circuit I at time t. Instantaneous voltage value output by bridge circuit II ;

[0032] (3) Calculate the predicted instantaneous voltage value at time t The residual of bridge I at time t The residual of bridge section II at time t The formula is as follows:

[0033] ;

[0034] ;

[0035] ;

[0036] In the formula, These are the preset fusion weight coefficients;

[0037] (4) Calculate the moving average of the voltage residual of bridge circuit I at time t. The moving average of the voltage residual of bridge circuit II at time t The formula is as follows:

[0038] ;

[0039] ;

[0040] In the formula, This is a historical weighting factor, with a value range of 0.85-0.95. Users can customize the specific value. The previous time step at time t , The previous time step at time t When t=0, let =0, let =0;

[0041] (5) Calculate the two instantaneous rates of change, respectively and When t=0, let =0, let =0;

[0042] (6) Determine whether conditions ① and ② are both true. Condition ① is: and Condition ② is: and , The set threshold;

[0043] If both conditions are met, it is determined that both Bridge I and Bridge II have experienced unidirectional drift, and the program terminates.

[0044] If only condition ① is met, it is determined that bridge road I has experienced unidirectional drift, and the measurement work is switched to bridge road II, and the program terminates;

[0045] If only condition ② is met, it is determined that bridge road II has experienced unidirectional drift, and the measurement work is switched to bridge road I, and the program terminates.

[0046] If none of them are true, proceed to step (7);

[0047] (7) Update sampling time: Let t = t + Δt, and return to step (2).

[0048] As described above, the fault diagnosis module of a miniature servo electric cylinder includes a third fault diagnosis module. This third module is used to detect oscillating drift faults in bridge circuit I and bridge circuit II. The detection principle is as follows: when oscillating drift occurs in the bridge circuit, the bridge voltage residual will deviate periodically, causing the average voltage residual to fluctuate around the overall average, but the baseline will rise. Therefore, whether the baseline rise of the average voltage residual exceeds a threshold can be used to determine whether oscillating drift has occurred in the bridge circuit. The specific workflow is as follows:

[0049] (1) Initialize parameters: Let t=0 (initial time), set the sampling time interval Δt (Δt=0.01s-0.1s, the specific value can be customized by the user, preferably 0.05s);

[0050] (2) Collect the instantaneous voltage value of bridge circuit I at time t. Instantaneous voltage value output by bridge circuit II ;

[0051] (3) Calculate the predicted instantaneous voltage value at time t The residual of bridge I at time t The residual of bridge section II at time t The formula is as follows:

[0052] ;

[0053] ;

[0054] ;

[0055] In the formula, These are the preset fusion weight coefficients;

[0056] (4) Calculate the moving average of the voltage residual of bridge circuit I at time t. The moving average of the voltage residual of bridge circuit II at time t The formula is as follows:

[0057] ;

[0058] ;

[0059] In the formula, This is a historical weighting factor, with a value range of 0.85-0.95. Users can customize the specific value. The previous time step at time t , The previous time step at time t When t=0, let =0, let =0;

[0060] (5) Calculate the baseline rise rate of bridge section I at time t. The baseline rise rate of bridge section II at time t The formula is as follows:

[0061] ;

[0062] ;

[0063] In the formula, Represents the current iteration number. Representing the During the next iteration The value (e.g., when Δt = 0.1s, in the first cycle) It is 0 in the second loop. (0.1, and so on). Representing the During the next iteration The value, Representing the During the next iteration The value of ; when t=0, let , ;

[0064] (6) Fault diagnosis:

[0065] judge and Whether they are true at the same time, The set threshold;

[0066] If both conditions are met, it is determined that both Bridge I and Bridge II have experienced oscillatory drift, and the program terminates.

[0067] If only If the condition is met, it is determined that bridge path I has experienced oscillating drift, and the measurement work is switched to bridge path II, and the program terminates.

[0068] If only If the condition is met, it is determined that bridge path II has experienced oscillating drift, and the measurement work is switched to bridge path I, and the program terminates.

[0069] If none of them are true, proceed to step (7);

[0070] (7) Update sampling time: Let t = t + Δt, and return to step (2).

[0071] As described above, a miniature servo electric cylinder, The acquisition process is as follows: During the initialization phase of the micro servo electric cylinder, n instantaneous voltage values ​​output by bridge circuit I are continuously acquired, and their standard deviation is calculated. Given n values ​​≤ 200, synchronously collect n instantaneous voltage values ​​output by bridge circuit II and calculate their standard deviation. Substitute into the following formula to calculate:

[0072] .

[0073] As described above, the fault diagnosis module of this miniature servo electric cylinder includes a fourth fault diagnosis module. This fourth fault diagnosis module is used to detect short-circuit faults in bridge circuit I and bridge circuit II. The detection principle is as follows: when a short-circuit fault occurs in the circuit, the resistance R of one arm of the bridge circuit changes from 0 to ∞. At this time, the common-mode voltage value of the two bridge circuits will change abruptly. To avoid the randomness caused by the noise change, this module uses the moving average calculation to make the judgment. The specific workflow is as follows:

[0074] (1) Initialize parameters: Let t=0 (initial time), set the sampling time interval Δt (Δt=0.01s-0.1s, the specific value can be customized by the user, preferably 0.05s);

[0075] (2) Collect the instantaneous voltage value of bridge circuit I at time t. Instantaneous voltage value output by bridge circuit II ;

[0076] (3) Calculate the common-mode voltage at time t. :

[0077] ;

[0078] (4) Calculate the moving average of the common-mode voltage of bridge circuit I at time t. The moving average of the common-mode voltage of bridge circuit II at time t The formula is as follows:

[0079] ;

[0080] ;

[0081] In the formula, This is a historical weighting factor, with a value range of 0.85-0.95. Users can customize the specific value. The previous time step at time t , The previous time step at time t When t=0, let =0, let =0;

[0082] (5) Judgment and Whether they are true at the same time, The set threshold;

[0083] If both conditions are met, then it is determined that both Bridge I and Bridge II are short-circuited, and the procedure terminates.

[0084] If only If the condition is met, it is determined that Bridge I is short-circuited, and the measurement work is switched to Bridge II, and the program terminates.

[0085] If only If the condition is met, it is determined that Bridge Circuit II is short-circuited, and the measurement work is switched to Bridge Circuit I, and the program terminates.

[0086] If none of them are true, proceed to step (6).

[0087] (6) Update sampling time: Let t = t + Δt, and return to step (2).

[0088] Common sensor faults include: open circuit (manifesting as a sudden drop in signal), short circuit (causing output drift), and drift (data jitter). Existing sensor redundancy fault detection algorithms employ the following methods to address these faults:

[0089] ① Multi-result comparison method: Existing technologies use voting or averaging strategies to output results, which cannot identify faults when both bridges drift simultaneously (such as when temperatures change synchronously). In addition, this method is mostly applicable to static or low-speed systems and is not suitable for measuring the force and fault tolerance of electric cylinder push rods during high-speed movement.

[0090] The multi-result comparison method of this invention abandons the voting mechanism and achieves independent analysis of dual signals through open circuit fault judgment (residual mutation detection), short circuit fault judgment (common mode voltage polarity identification), and drift fault judgment (rate of change and baseline rise analysis).

[0091] For example, in wire breakage detection, when and Simultaneously, the dual-bridge fault is immediately identified upon establishment, completely avoiding the risk of the fault being masked.

[0092] ② Inverse voltage calculation method: The traditional inverse voltage calculation method relies on a relatively fragile circuit model and has a high false alarm rate under dynamic interference.

[0093] This invention innovatively uses data fusion to predict values. As a dynamic benchmark, the weighting coefficient K is calibrated using real-time variance ( This design eliminates the need for a fixed model, directly generating theoretical values ​​through cross-calibration of data from two bridges, thus reducing the error amplification problem caused by drift interference at the source.

[0094] ③ Support Vector Machine (SVM) method: The SVM method requires a large amount of labeled fault data and has high prediction latency, making it difficult to run in real time. All fault detection modules in this invention use unsupervised recursive algorithms: the moving average calculation only requires historical state caching.

[0095] This invention employs a dual-bridge circuit operating simultaneously and uses data fusion to predict and smooth the output of the dual-bridge voltage, thereby improving output accuracy.

[0096] This invention integrates a triple fault protection mechanism of open circuit, short circuit, and drift, and achieves automatic disconnection of faulty bridge circuits and seamless switching of single bridges through redundant design:

[0097] 1) Wire breakage fault identification is achieved through dual-bridge residual mutation detection. .

[0098] 2) Accurately determine short circuits by identifying common-mode voltage polarity. .

[0099] 3) Drift faults are captured and identified by combining the rate of change threshold and the baseline rise rate, which extends the service life of the electric cylinder sensor force control and avoids equipment downtime due to single faults.

[0100] Furthermore, the recursive computation characteristics of data fusion enable the algorithm to run on low-end MCUs, reducing power consumption, and unlike support vector methods, it does not require external training data or complex model calibration.

[0101] High-precision applications of miniature servo electric cylinders (such as humanoid robot dexterous hands, medical devices, and microelectronic press fitting) place extremely high demands on force control reliability: dexterous hands rely on force feedback to control gripping force, and sensor failure may lead to gripping failure or crushing of objects; in medical devices, force errors exceeding 0.5N may damage the patient's nerves; in microelectronic press fitting, force feedback failure can cause process stagnation or damage to brittle materials. The redundant design and fault diagnosis mechanism of this invention can effectively avoid the above risks.

[0102] As described above, in a miniature servo electric cylinder, the sensor flange disc is composed of a ring and a crossbeam-type elastic body located inside the ring;

[0103] The cross-beam type elastomer is a one-piece molded part;

[0104] The cross-beam type elastic body consists of a load-bearing beam, four deformable beams of equal length, and four hollow beams of equal length. The central axis of the load-bearing beam coincides with the central axis of the ring. The deformable beams and hollow beams are arranged radially along the ring. One end of each of the four deformable beams is connected to the load-bearing beam, and the other end is connected to one end of each of the four hollow beams. The other end of each of the four hollow beams is connected to the inner wall of the ring.

[0105] Four strain gauges are respectively attached to the upper surface of the four deformable beams and are evenly distributed in a 90° ring around the stressed beam.

[0106] Four second strain gauges are respectively attached to the lower surface of the four deformable beams and are symmetrically distributed with the four first strain gauges;

[0107] Both the first and second strain gauges reflect the pushing and pulling forces acting on the push rod by detecting the strain of the deformed beam.

[0108] This invention improves force control accuracy through symmetrical patching and structural optimization: Under load, the push-pull force of the push rod causes the load-bearing beam to deform, causing the deformed beam to bend upwards (push force) or downwards (pull force), thereby changing the resistance and output voltage of the strain gauges. The force value can be obtained through calibration. The hollow beam design enhances deformation sensitivity, and its material can be adjusted as needed to adapt to different force ranges; the average voltage of the strain gauges on both sides can be used as the output, and faults can be identified through cross-calibration.

[0109] Patent CN117906792B discloses a tension and compression sensor. The sensor adopts a split structure design. Component 21 is the force-bearing component, and tension and compression are applied to 21 through the cooperation of component 15 and component 35. Component 22 is a cross spring force-bearing component, and its deformation is detected by strain gauges. The original patent description states that "based on the resistance strain effect and converting mechanical strain into resistance change, the spring steel force-bearing component 22 (i.e., strain detection carrier) is mainly composed of a sensitive grid (usually made of fine metal wires or metal foil), a substrate, and leads."

[0110] According to formula 1 for the output voltage of the Wheatstone bridge ( When the strain gauge is not deformed by external force, if R1R3=R2R4, the bridge is in equilibrium, and the output voltage is 0. However, when the strain gauge is deformed by external force, the changes in the four resistors must be symmetrically distributed to maintain consistency (as shown by ΔR in Formula 1). However, the integral strain gauge used in patent CN117906792B, under unilateral tension or compression, has a non-symmetrical distribution of internal forces, resulting in different changes in the ΔR of each resistor. This makes it impossible to completely eliminate the ΔR term in the denominator of Formula 1, causing the denominator to exhibit non-linear characteristics, leading to voltage fluctuations and reduced measurement accuracy. Furthermore, this asymmetrical bonding method can cause inconsistent strain transmission paths, leading to localized stress concentration and measurement errors.

[0111] To address the above deficiencies, this invention employs a top-to-bottom symmetrical patch method, such as... Figure 6 As shown, Figure 6 The four resistors are R1, R2, R3, and R4. When attaching the strain gauges, their positions are as follows: Figure 6 As shown, R1 and R2 are symmetrically attached to both sides of the load-bearing beam, and R3 and R4 are symmetrically attached to both sides of the load-bearing beam. Compared with existing technology, this ensures the consistency of the resistance deformation of the two sets of strain gauges and the oppositeity of the deformation directions during deformation. Specifically, the force directions of R1 and R2 are the same, the force directions of R3 and R4 are the same, the force directions of R1 and R4 are opposite, and the force directions of R2 and R3 are opposite, which can better satisfy Formula 2 (substituting into the general formula: After simplification, we get Formula 2: To meet the requirements, the ΔR term in the denominator should be eliminated to the greatest extent possible, thus satisfying the final result of Formula 2. This ensures that the output voltage change ΔU is only affected by the change in ΔR, where U is the output voltage (constant) and R is the initial resistance of the strain gauge (constant). ΔU and ΔR are directly proportional. Symmetrical bonding avoids inconsistent strain transmission paths, reduces errors caused by local stress concentration, and thus provides greater voltage output accuracy.

[0112] As described above, in a miniature servo electric cylinder, the upper surface of the force-bearing beam is flat, and a threaded rod is fixed on the lower surface of the force-bearing beam. The threaded rod is used to connect with the workpiece when the miniature servo electric cylinder is working.

[0113] The miniature servo electric cylinder described above also includes a gearbox and a trapezoidal screw; the sensor flange, hollow cup motor, gearbox, and trapezoidal screw are arranged from bottom to top;

[0114] The gearbox housing contains an internal gear ring, planetary gears, a sun gear, and a planetary gear carrier. The sun gear meshes with the planetary gears, the planetary gears mesh with the internal gear ring, and the planetary gear carrier connects the planetary gears and the sun gear in series.

[0115] The trapezoidal screw includes a flange section, a stepped optical shaft section, and a threaded section arranged from bottom to top;

[0116] The output shaft of the coreless motor is coaxial with the sun gear and the two are fixedly connected; the flange part of the trapezoidal screw is connected to the planetary gear; the push rod has a cylindrical structure, and the push rod is sleeved on the threaded part of the trapezoidal screw and threadedly connected to it; an anti-rotation slider is installed at the lower end of the push rod.

[0117] As described above, in a miniature servo electric cylinder, a first position sensor and a second position sensor are embedded in slots on both sides of the anti-rotation slider.

[0118] Traditional position feedback relies on a single encoder or potentiometer, which has significant drawbacks: a single encoder cannot compensate for mechanical backlash, and gear machining errors can amplify deviations; a single potentiometer is prone to contact surface wear due to pushrod off-center loading or vibration, causing resistance jumps or signal drift, resulting in low reliability in high-dynamic scenarios such as medical equipment. This invention addresses this by placing a first position sensor and a second position sensor on both sides of the anti-rotation slider. The output signals from the first and second position sensors are fused using an averaging algorithm to reduce yaw error; the signal difference between the two sensors is used to compensate for transmission chain backlash in real time.

[0119] As described above, in a miniature servo electric cylinder, a thrust bearing is fitted onto the stepped optical shaft portion of the trapezoidal screw, and the inner ring of the thrust bearing is interference-fitted with the stepped optical shaft portion of the trapezoidal screw.

[0120] The miniature servo electric cylinder described above also includes a third end cap, a second housing, a first housing, a second end cap, and a first end cap;

[0121] The third end cover, the second housing, the gearbox, the first housing, the second end cover, the sensor flange, and the first end cover are connected in sequence from top to bottom;

[0122] The first end cap is fixedly connected to the sensor flange disc body by bolts; the second end cap is connected to the first housing by bolts; the third end cap constrains the stroke of the push rod through a limiting structure.

[0123] The trapezoidal screw, thrust bearing, and push rod are all located inside the second housing. The interior of the second housing is provided with a circular groove and a square slide. The outer ring of the thrust bearing is embedded in the circular groove, the push rod is slidably connected to the square slide, and the anti-rotation slider is stuck in the square slide.

[0124] The main body of the coreless motor is installed inside the first housing.

[0125] Traditional electric cylinders typically feature a separate design for the gearbox and main housing, connected by flanges or bolts. This results in a large overall size, making them unsuitable for confined spaces such as those occupied by dexterity devices. Furthermore, the internal interconnection relies on flexible circuit boards or flat cables, which are prone to fatigue and breakage during high-frequency extension and retraction. Additionally, redundant wiring space is required, conflicting with miniaturization goals. This invention integrates the third end cover, second housing, and gearbox into a single unit, eliminating separate assembly interfaces and flat cables, minimizing space requirements, and simultaneously improving structural stability.

[0126] The reason why this invention can achieve the effect that both the first strain gauge and the second strain gauge reflect the pushing and pulling force on the push rod by detecting the strain of the deformed beam is because:

[0127] ① The first end cap is fixedly connected to the ring of the sensor flange by bolts to form a closed load-bearing structure; the second end cap is connected to the first housing by bolts to constrain the axial displacement of the hollow cup motor; together they ensure that the push and pull force is transmitted to the cross beam elastic body without loss.

[0128] ② The push rod is connected to the planetary gear through the trapezoidal screw. The planetary gear meshes with the sun gear. The reduction gearbox is a three-stage reduction mechanism. The first-stage sun gear of the reduction gearbox is matched with the output shaft of the hollow cup motor, and finally the load is transmitted to the force beam of the sensor flange. In this path: the stepped optical shaft part of the trapezoidal screw is interference-fitted with the inner ring of the thrust bearing. The outer ring of the thrust bearing is embedded in the circular groove of the second housing. The anti-rotation slider is stuck in the square slide of the second housing.

[0129] ③ The crossbeam-type elastomer of the sensor flange is machined as a single piece (non-welded) to eliminate assembly stress; the first housing covers the hollow cup motor to isolate electromagnetic interference; the third end cover constrains the push rod stroke through a limiting structure to prevent overload impact.

[0130] Beneficial effects:

[0131] (1) The present invention solves the problem of force control mode failure caused by single-surface strain gauge failure in the existing single force sensor configuration. By setting the first strain gauge and the second strain gauge on the upper and lower surfaces of the sensor flange respectively, a dual-bridge redundant structure of bridge circuit I and bridge circuit II is formed. When one bridge circuit fails, it can switch to the other bridge circuit to continue working, maintain the continuous operation of the force control function, and significantly improve the system reliability.

[0132] (2) The present invention detects the open circuit, short circuit, unidirectional drift and oscillating drift faults of the dual bridge circuit through the fault judgment module. When both bridge circuits are faulty, the shutdown protection is triggered, which avoids the irreversible result caused by the fault applying the wrong force value to the target object, and improves the safety of the micro servo electric cylinder in high precision application scenarios.

[0133] (3) The voltage signal output by the dual bridge circuit is fused by Kalman filtering to obtain the real-time force value. Combined with the position information and signal difference output by the first position sensor and the second position sensor through the mean value algorithm to compensate for the transmission gap, high-precision closed-loop control is realized, which improves the accuracy of force control and position control.

[0134] (4) The cross-beam type elastomer of the sensor flange adopts an integral molding design. The first strain gauge and the second strain gauge are symmetrically attached to the upper and lower surfaces of the deformable beam, which reduces the measurement error caused by inconsistent strain transmission paths and local stress concentration, and further improves the force measurement accuracy.

[0135] (5) The present invention eliminates the separate assembly interface and flexible cable by integrating the third end cover, the second housing, the gearbox and other components, thus realizing miniaturized design and adapting to application scenarios in narrow spaces. Attached Figure Description

[0136] Figure 1 and Figure 2 This is a schematic diagram of the structure of the miniature servo electric cylinder of the present invention; wherein, Figure 1 For assembly diagram, Figure 2 This is an exploded view;

[0137] Figure 3 This is a schematic diagram of the sensor flange structure; where (a) is viewed from the lower side and (b) is viewed from the upper side.

[0138] Figure 4 This is a structural schematic diagram of the gearbox; where (a) is an assembly diagram and (b) is an exploded view.

[0139] Figure 5 This is a schematic diagram showing the connection relationship between the push rod, the anti-rotation slider, the first position sensor, and the second position sensor; where (a) is a side view and (b) is a top view.

[0140] Figures 1-5In the diagram, 1-first end cap, 2-sensor flange, 2a-force beam, 2b-deformable beam, 2c-hollow beam, 2d-first strain gauge, 2e-second strain gauge, 3-second end cap, 4-hollow cup motor, 5-first housing, 6-reduction gearbox, 6a-planetary gear, 6b-sun gear, 6c-planetary gear carrier, 7-trapezoidal screw, 8-thrust bearing, 9-push rod, 9a-anti-rotation slider, 9b-first position sensor, 9c-second position sensor, 10-second housing, 11-third end cap;

[0141] Figure 6 The diagram shows the layout and stress of a symmetrical patch strain gauge; (a) is viewed from the lower side, and (b) is viewed from the upper side. Detailed Implementation

[0142] The present invention will be further described below with reference to specific embodiments. It should be understood that these embodiments are for illustrative purposes only and are not intended to limit the scope of the invention. Furthermore, it should be understood that after reading the teachings of this invention, those skilled in the art can make various alterations or modifications to the invention, and these equivalent forms also fall within the scope defined by the appended claims.

[0143] A miniature servo electric cylinder, such as Figure 1 and Figure 2 As shown, it includes a fault diagnosis module, a push rod 9, a hollow cup motor 4, a sensor flange 2, a gearbox 6, a trapezoidal screw 7, a third end cover 11, a second housing 10, a first housing 5, a second end cover 3, and a first end cover 1;

[0144] The third end cover 11, the second housing 10, the gearbox 6, the first housing 5, the second end cover 3, the sensor flange 2, and the first end cover 1 are connected in sequence from top to bottom;

[0145] like Figure 3As shown, the sensor flange 2 includes a horizontally arranged disc, which consists of a ring and a cross-beam type elastic body located within the ring. The cross-beam type elastic body is a one-piece molded component, consisting of a load-bearing beam 2a, four deformable beams 2b of equal length, and four hollow beams 2c of equal length. The central axis of the load-bearing beam 2a coincides with the central axis of the ring. The deformable beams 2b and the hollow beams 2c are arranged radially along the ring. One end of each of the four deformable beams 2b is connected to the load-bearing beam 2a, and the other end is connected to one end of each of the four hollow beams 2c. The other end of each of the four hollow beams 2c is connected to the inner wall of the ring. The upper surface of the load-bearing beam 2a is flat, and a threaded rod is fixed on the lower surface of the load-bearing beam 2a. The threaded rod is used for micro servo... When the electric cylinder is working, it is connected to the workpiece; a first strain gauge 2d is attached to the upper surface of each of the four deformable beams 2b, and each first strain gauge 2d is evenly distributed in a 90° ring around the force beam 2a. Each first strain gauge 2d is connected through a resistor and an amplifier to form bridge circuit I; a second strain gauge 2e is attached to the lower surface of each of the four deformable beams 2b, and each second strain gauge 2e is symmetrically distributed with each first strain gauge 2d. Each second strain gauge 2e is connected through a resistor and an amplifier to form bridge circuit II; both bridge circuit I and bridge circuit II are Wheatstone bridges with full bridge connection. The first strain gauge 2d and the second strain gauge 2e reflect the push-pull force on the push rod 9 by detecting the strain of the deformable beam 2b;

[0146] like Figure 1 and Figure 2 As shown, the first end cap 1 is fixedly connected to the disc body of the sensor flange 2 by bolts; the second end cap 3 is connected to the first housing 5 by bolts; the main body of the hollow cup motor 4 is installed inside the first housing 5, and the output shaft of the hollow cup motor 4 is parallel to the vertical direction;

[0147] like Figure 4 As shown, the gearbox 6 is equipped with an internal gear ring, planetary gear 6a, sun gear 6b ​​and planetary gear carrier 6c. The sun gear 6b ​​meshes with the planetary gear 6a, the planetary gear 6a meshes with the internal gear ring, and the planetary gear carrier 6c connects the planetary gear 6a and the sun gear 6b ​​in series. The sun gear 6b ​​is coaxial with the output shaft of the hollow cup motor 4 and the two are fixedly connected.

[0148] like Figure 2 As shown, the trapezoidal screw 7 includes a flange portion, a stepped optical shaft portion, and a threaded portion arranged sequentially from bottom to top; the flange portion of the trapezoidal screw 7 is connected to the planetary gear 6a; a thrust bearing 8 is fitted onto the stepped optical shaft portion of the trapezoidal screw 7, and the inner ring of the thrust bearing 8 is interference-fitted with the stepped optical shaft portion of the trapezoidal screw 7; a push rod 9 is parallel to the vertical direction, and the push rod 9 has a cylindrical structure, fitting onto and threadedly connecting to the threaded portion of the trapezoidal screw 7; as shown... Figure 5 As shown, an anti-rotation slider 9a is installed at the lower end of the push rod 9, and a first position sensor 9b and a second position sensor 9c are embedded in slots on both sides of the anti-rotation slider 9a.

[0149] like Figure 1 and Figure 2 As shown, the trapezoidal screw 7, the thrust bearing 8, and the push rod 9 are all located inside the second housing 10. The interior of the second housing 10 is provided with a circular groove and a square slide. The outer ring of the thrust bearing 8 is embedded in the circular groove, the push rod 9 is slidably connected to the square slide, and the anti-rotation slider 9a is inserted into the square slide.

[0150] The third end cap 11 constrains the stroke of the push rod 9 through a limiting structure.

[0151] The fault diagnosis module includes a first fault diagnosis module, a second fault diagnosis module, a third fault diagnosis module, and a fourth fault diagnosis module;

[0152] The first fault diagnosis module is used to detect open circuit faults in Bridge Circuit I and Bridge Circuit II. The specific workflow is as follows:

[0153] (1) Initialize parameters: Set t=0 and set the sampling time interval Δt;

[0154] (2) Collect the instantaneous voltage value of bridge circuit I at time t. Instantaneous voltage value output by bridge circuit II ;

[0155] (3) Calculate the predicted instantaneous voltage value at time t The residual of bridge I at time t The residual of bridge section II at time t The formula is as follows:

[0156] ;

[0157] ;

[0158] ;

[0159] In the formula, The preset fusion weight coefficients, The acquisition process is as follows: During the initialization phase of the micro servo electric cylinder, n instantaneous voltage values ​​output by bridge circuit I are continuously acquired, and their standard deviation is calculated. Given n values ​​≤ 200, synchronously collect n instantaneous voltage values ​​output by bridge circuit II and calculate their standard deviation. Substitute into the following formula to calculate:

[0160] ;

[0161] (4) Determine whether conditions ① and ② are both true. Condition ① is: and Condition ② is: and ;

[0162] If both conditions are met, then it is determined that both Bridge Road I and Bridge Road II are disconnected, and the program terminates.

[0163] If only condition ① is met, then it is determined that bridge road I is disconnected, and the measurement work is switched to bridge road II, and the program terminates;

[0164] If only condition ② is met, then it is determined that bridge road II is disconnected, and the measurement work is switched to bridge road I, and the program terminates;

[0165] If none of them are true, proceed to step (5);

[0166] (5) Update sampling time: Let t = t + Δt, and return to step (2).

[0167] The second fault diagnosis module is used to detect unidirectional drift faults in bridge circuit I and bridge circuit II. The specific workflow is as follows:

[0168] (1) Initialize parameters: Set t=0 and set the sampling time interval Δt;

[0169] (2) Collect the instantaneous voltage value of bridge circuit I at time t. Instantaneous voltage value output by bridge circuit II ;

[0170] (3) Calculate the predicted instantaneous voltage value at time t The residual of bridge I at time t The residual of bridge section II at time t The formula is as follows:

[0171] ;

[0172] ;

[0173] ;

[0174] In the formula, The preset fusion weight coefficients, The acquisition process is as follows: During the initialization phase of the micro servo electric cylinder, n instantaneous voltage values ​​output by bridge circuit I are continuously acquired, and their standard deviation is calculated. Given n values ​​≤ 200, synchronously collect n instantaneous voltage values ​​output by bridge circuit II and calculate their standard deviation. Substitute into the following formula to calculate:

[0175] ;

[0176] (4) Calculate the moving average of the voltage residual of bridge circuit I at time t. The moving average of the voltage residual of bridge circuit II at time t The formula is as follows:

[0177] ;

[0178] ;

[0179] In the formula, This is a historical weighting factor, with a value range of 0.85-0.95; The previous time step at time t , The previous time step at time t When t=0, let =0, let =0;

[0180] (5) Calculate the two instantaneous rates of change, respectively and When t=0, let =0, let =0;

[0181] (6) Determine whether conditions ① and ② are both true. Condition ① is: and Condition ② is: and , The set threshold;

[0182] If both conditions are met, it is determined that both Bridge I and Bridge II have experienced unidirectional drift, and the program terminates.

[0183] If only condition ① is met, it is determined that bridge road I has experienced unidirectional drift, and the measurement work is switched to bridge road II, and the program terminates;

[0184] If only condition ② is met, it is determined that bridge road II has experienced unidirectional drift, and the measurement work is switched to bridge road I, and the program terminates.

[0185] If none of them are true, proceed to step (7);

[0186] (7) Update sampling time: Let t = t + Δt, and return to step (2).

[0187] The third fault diagnosis module is used to detect oscillating drift faults in bridge circuit I and bridge circuit II. The specific workflow is as follows:

[0188] (1) Initialize parameters: Set t=0 and set the sampling time interval Δt;

[0189] (2) Collect the instantaneous voltage value of bridge circuit I at time t. Instantaneous voltage value output by bridge circuit II ;

[0190] (3) Calculate the predicted instantaneous voltage value at time t The residual of bridge I at time t The residual of bridge section II at time t The formula is as follows:

[0191] ;

[0192] ;

[0193] ;

[0194] In the formula, The preset fusion weight coefficients, The acquisition process is as follows: During the initialization phase of the micro servo electric cylinder, n instantaneous voltage values ​​output by bridge circuit I are continuously acquired, and their standard deviation is calculated. Given n values ​​≤ 200, synchronously collect n instantaneous voltage values ​​output by bridge circuit II and calculate their standard deviation. Substitute into the following formula to calculate:

[0195] ;

[0196] (4) Calculate the moving average of the voltage residual of bridge circuit I at time t. The moving average of the voltage residual of bridge circuit II at time t The formula is as follows:

[0197] ;

[0198] ;

[0199] In the formula, This is a historical weighting factor, with a value range of 0.85-0.95; The previous time step at time t , The previous time step at time t When t=0, let =0, let =0;

[0200] (5) Calculate the baseline rise rate of bridge section I at time t. The baseline rise rate of bridge section II at time t The formula is as follows:

[0201] ;

[0202] ;

[0203] In the formula, Represents the current iteration number. Representing the During the next iteration The value, Representing the During the next iteration The value, Representing the During the next iteration The value of ; when t=0, let , ;

[0204] (6) Fault diagnosis:

[0205] judge and Whether they are true at the same time, The set threshold;

[0206] If both conditions are met, it is determined that both Bridge I and Bridge II have experienced oscillatory drift, and the program terminates.

[0207] If only If the condition is met, it is determined that bridge path I has experienced oscillating drift, and the measurement work is switched to bridge path II, and the program terminates.

[0208] If only If the condition is met, it is determined that bridge path II has experienced oscillating drift, and the measurement work is switched to bridge path I, and the program terminates.

[0209] If none of them are true, proceed to step (7);

[0210] (7) Update sampling time: Let t = t + Δt, and return to step (2).

[0211] The fourth fault diagnosis module is used to detect short-circuit faults in bridge circuit I and bridge circuit II. The specific workflow is as follows:

[0212] (1) Initialize parameters: Set t=0 and set the sampling time interval Δt;

[0213] (2) Collect the instantaneous voltage value of bridge circuit I at time t. Instantaneous voltage value output by bridge circuit II ;

[0214] (3) Calculate the common-mode voltage at time t. :

[0215] ;

[0216] (4) Calculate the moving average of the common-mode voltage of bridge circuit I at time t. The moving average of the common-mode voltage of bridge circuit II at time t The formula is as follows:

[0217] ;

[0218] ;

[0219] In the formula, This is a historical weighting factor, with a value range of 0.85-0.95; The previous time step at time t , The previous time step at time t When t=0, let =0, let =0;

[0220] (5) Judgment and Whether they are true at the same time, The set threshold;

[0221] If both conditions are met, then it is determined that both Bridge I and Bridge II are short-circuited, and the procedure terminates.

[0222] If only If the condition is met, it is determined that Bridge I is short-circuited, and the measurement work is switched to Bridge II, and the program terminates.

[0223] If only If the condition is met, it is determined that Bridge Circuit II is short-circuited, and the measurement work is switched to Bridge Circuit I, and the program terminates.

[0224] If none of them are true, proceed to step (6).

[0225] (6) Update sampling time: Let t = t + Δt, and return to step (2).

[0226] When using this miniature servo electric cylinder, firstly, the threaded rod on the lower surface of the load-bearing beam is connected to the target workpiece. After power is applied, the hollow cup motor starts, and its output shaft drives the sun gear to rotate. After reduction by the planetary gears, it drives the trapezoidal screw to rotate. Through the threaded engagement, it pushes the push rod to move linearly along the square slide. The anti-rotation slider is locked in the slide to prevent the push rod from rotating with the screw, and the thrust bearing bears the axial force to ensure transmission stability. During operation, the deformable beam bends as the push rod is subjected to force. The first strain gauge and the second strain gauge output voltage signals through bridge circuit I and bridge circuit II, respectively. After dual-bridge data fusion, the real-time force value is obtained. The first position sensor and the second position sensor output position information through the averaging algorithm, and at the same time, the transmission gap is compensated by the signal difference to achieve high-precision closed-loop control. The fault judgment module monitors the status of the dual bridge circuits throughout the process. If a disconnection, short circuit, or drift (unidirectional drift, oscillating drift) is detected, the system automatically switches to normal bridge circuit operation; if both bridge circuits are faulty, a shutdown protection is triggered.

[0227] To demonstrate the effectiveness of fault detection, stability of force control accuracy, and miniaturization adaptability of the dual-bridge redundant design in this invention, and to verify the detection response speed and fault-tolerant switching performance of four types of faults—open circuit, short circuit, unidirectional drift, and oscillating drift—while quantifying force control error and position control accuracy, this invention is verified through the following experiments:

[0228] I. Wire breakage fault detection experiment;

[0229] Experimental principle: When a break-through fault occurs in bridge circuit I or bridge circuit II, the output voltage of the corresponding bridge circuit will suddenly drop to 0. According to the formula... , , If the residual value exceeds the set threshold If the corresponding bridge output voltage is <0.1V, a fault determination is triggered; the system needs to automatically identify the faulty bridge and switch to the normal bridge to maintain the continuous operation of the force control function.

[0230] Experimental equipment: as described above, a miniature servo electric cylinder, a precision tooling fixture (levelness error <0.1°), a 10N standard weight (accuracy ±0.01%), a dual-channel signal cut-off device (response time <1ms), and a high-speed data logger (sampling rate 10kHz).

[0231] The experimental steps are as follows:

[0232] (1) Prototype installation: The miniature servo electric cylinder is vertically fixed to the precision tooling, a matching force sensor is installed at the top of the push rod, a 10N standard weight load is applied, the force control mode is started and the 10N output is maintained;

[0233] (2) Parameter initialization: Set Δt=0.05s; During the initialization phase of the electric cylinder, continuously collect n=100 instantaneous voltage values ​​of bridge circuit I to calculate the standard deviation σ1, and simultaneously collect 100 instantaneous voltage values ​​of bridge circuit II to calculate the standard deviation σ2, and substitute them into the fusion weighting coefficient formula. Calculations yielded ;

[0234] (3) Fault injection and monitoring: Maintain stable system operation during the 0-5s phase and monitor and Synchronization; at t=5.000s, the signal line of bridge circuit I is cut off by the dual-channel signal cutter, injecting a line break fault; continuous monitoring continues for 5s. The sudden change process, the fault judgment delay time, and the force control output stability of bridge circuit II after switching;

[0235] The experimental results are as follows:

[0236] (a) Voltage and residual abrupt response (see Table 1);

[0237] Table 1

[0238]

[0239] (b) Fault diagnosis and switchover verification;

[0240] Residual calculation: At t=5.000s, the result is obtained according to the formula. ;

[0241] Threshold comparison: Based on the judgment criteria, 0.7× ;

[0242] Judgment result: 0.479V > 0.349V, and This meets the criteria for determining a broken line fault in Bridge Road I.

[0243] Switching action: At t=5.005s, the system automatically activates the independent output of Bridge Circuit II, and the force control mode is not interrupted;

[0244] (c) Accuracy verification (see Table 2);

[0245] Table 2

[0246] parameter Theoretical design value Measured value error Fault determination delay ≤0.05s 0.005s -90% Force control fluctuation <±0.5N ±0.18N Meets standards False positive rate 0% 0 / 20 times Meets standards

[0247] Experimental Conclusion: This experiment accurately reproduces the logic for detecting open circuit faults, and the system can pass the test. , With 0.7 The threshold comparison allows for rapid identification of bridge circuit disconnection faults. The sub-millisecond (5ms) response speed and seamless switching with the bridge circuit verify the effectiveness of the first fault judgment module and meet the requirement of "fault-free force control" in high-precision scenarios.

[0248] II. Short-circuit fault detection experiment;

[0249] Experimental principle: When a short-circuit fault occurs in bridge circuit I or bridge circuit II, the resistance of one arm of the corresponding bridge circuit approaches 0, resulting in a change in the common-mode voltage of the two bridge circuits. A sudden change occurs; based on the short-circuit fault detection logic, calculate... , ,like or If so, a short-circuit fault determination is triggered and bridge switching is completed;

[0250] Experimental equipment: as described above, a miniature servo electric cylinder, a programmable load controller (accuracy ±0.1N), a short-circuit fault simulator (with a 10KΩ parallel resistor), a high-precision data acquisition system (sampling rate 10kHz), and a constant temperature environment chamber (temperature control ±0.5℃).

[0251] The experimental steps are as follows:

[0252] (1) Prototype installation: The miniature servo electric cylinder is horizontally installed in a constant temperature environment chamber (set temperature 25℃), and the end of the push rod is connected to a programmable load controller to apply a constant load of 5N;

[0253] (2) Parameter initialization: Set Δt = 0.05s, =0.9, =0.1V; initialized at t=0. =0、 =0;

[0254] (3) Fault injection and monitoring: Maintain stable system operation during the 0-10s phase and record... Baseline value; at t=10.000s, a short-circuit fault (parallel 10KΩ resistor) is injected into bridge circuit I using a short-circuit fault simulator; continuous monitoring is performed after the short circuit. Mutation process, and The deviation changes are recorded, the fault judgment results and bridge switching timing are recorded, and when a short circuit in the dual bridge is triggered synchronously, the system judges the fault and terminates the program within 0.007s.

[0255] The experimental results are as follows:

[0256] (a) Common-mode voltage and moving mean deviation response (see Table 3);

[0257] Table 3

[0258]

[0259] (b) Fault diagnosis and switching;

[0260] Short circuit characteristics: Bridge circuit I voltage V1 jumps by 400% compared to normal conditions. Synchronous mutation;

[0261] Deviation calculation: at t=10.000s, =0.635V;

[0262] Threshold comparison: 0.635V > 0.1V ( (), which meets the criteria for determining a short-circuit fault in Bridge-Road I;

[0263] Switching action: At t=10.005s, the system automatically activates the independent output of bridge circuit II, and the force control is not interrupted;

[0264] (c) Accuracy verification (see Table 4);

[0265] Table 4

[0266] parameter Theoretical design value Measured value error Single short-circuit determination delay ≤0.05s 0.005s -90% Double short-circuit response ≤0.07s 0.007s -90% Force control fluctuation <±0.5N ±0.22N -56% False positive rate 0% 0 / 20 times 0%

[0267] Experimental conclusion: In the short-circuit fault detection experiment, the system can accurately identify short-circuit faults through common-mode voltage mutation and sliding mean deviation. The 5ms response speed is 10 times better than the design specifications. The force control accuracy is stable after switching, which verifies the reliability of the fourth fault judgment module.

[0268] III. One-way drift fault detection experiment;

[0269] Experimental principle: When a unidirectional drift fault occurs in bridge circuit I or bridge circuit II, the resistance of the corresponding bridge arm increases non-linearly. According to the unidirectional drift detection logic, , It will increase in one direction; through calculation , , , If condition ① or condition ② is met, a one-way drift judgment is triggered and the bridge path is switched.

[0270] Experimental equipment: as described above, a miniature servo electric cylinder, a high-resolution thermal imager (spatial resolution 0.1 mm), a dynamic signal analyzer (bandwidth 100 kHz), and a constant temperature environment chamber (temperature control range -20℃~150℃, accuracy ±0.5℃).

[0271] The experimental steps are as follows:

[0272] (1) Prototype pretreatment: The micro servo electric cylinder is vertically installed in a constant temperature environment chamber (initial temperature 25℃), the push rod applies a constant load of 8N, and the thermal imager is aimed at the strain gauge area to monitor the temperature distribution uniformity;

[0273] (2) Parameter initialization: Set the sampling time interval Δt = 0.05s. =0.9, =0.3V, the heating rate of the constant temperature chamber is set to 2℃ / min;

[0274] (3) Temperature drift simulation and monitoring: Temperature ranged from 25℃ to 35℃ during the 0-5 min period, and the temperature was recorded. During the 5-15 minute phase, the temperature changes from 35℃ to 55℃. Calculations are performed. After 15 minutes, the temperature will change from 55℃ to 70℃. Monitor whether the fault judgment conditions are met and the bridge switching status.

[0275] The experimental results are as follows:

[0276] (a) Temperature drift characteristics data (see Table 5);

[0277] Table 5

[0278]

[0279] (b) Fault diagnosis and switching;

[0280] At t=14.5min, =0.32V>0.3V ( ),and =0.021V / s>0, indicating that bridge path I has experienced unidirectional drift; at t=15.0min, the system automatically switches to bridge path II.

[0281] (c) Accuracy verification (see Table 6);

[0282] Table 6

[0283] parameter Theoretical design value Measured value error Fault determination delay ≤60s 28s -53% Voltage-temperature sensitivity 0.5V / 10℃ 0.48V / 10℃ -4% False positive rate ≤5% 1 / 20 times -80%

[0284] Experimental conclusion: In the unidirectional drift fault detection experiment, the system can accurately identify unidirectional drift caused by temperature drift through the dual conditions of "sliding mean + rate of change". After switching to bridge circuit II, the force control fluctuation is maintained at ±0.25N, which verifies the detection capability of the second fault judgment module for slow drift faults.

[0285] IV. Experiment on Detection of Oscillation-Type Drift Fault;

[0286] Experimental principle: When an oscillating drift fault occurs in a circuit, the voltage residual will fluctuate periodically and the baseline will continuously rise; through calculation... , ,like or This triggers an oscillating drift detection and switches the bridge path;

[0287] Experimental equipment: as described above, a miniature servo electric cylinder, a capacitive coupling interference device (adjustable from 10nF to 100nF), an alternating load generator (adjustable from 0.1 to 100Hz), a dynamic signal analyzer (sampling rate 100kHz), and a spectrum analyzer (resolution 1Hz).

[0288] The experimental steps are as follows:

[0289] (1) Prototype configuration: The miniature servo electric cylinder is installed on the vibration isolation platform, the push rod is connected to the alternating load generator and a 12N alternating load is applied; a 15nF capacitor is connected in parallel at the feedback end of the signal amplifier to induce the circuit to self-excited oscillation;

[0290] (2) Parameter initialization: Set the sampling time interval Δt = 0.05s. =0.9, =0.2, data calculation window r=10 (corresponding to 0.5s);

[0291] (3) Oscillation simulation and monitoring: Maintain system stability during the 0-10s phase and record the background noise and Baseline; 10-20s phase monitoring The fluctuation frequency and The baseline rise trend is calculated in real time. ;Continuously monitor the fault determination results and bridge switching effect after 20 seconds;

[0292] The experimental results are as follows:

[0293] (a) Oscillation characteristic data (see Table 7);

[0294] Table 7

[0295]

[0296] (b) Fault diagnosis and switching;

[0297] At t=20.0s, =0.29>0.2 ( Spectrum analysis showed a main oscillation frequency of 125Hz, indicating that bridge circuit I was experiencing oscillatory drift; the system immediately switched to bridge circuit II, and the oscillation signal disappeared;

[0298] (c) Accuracy verification (see Table 8);

[0299] Table 8

[0300] parameter Theoretical design value Measured value error Fault determination delay ≤5s 2.1s -58% Frequency resolution ≥50Hz 1Hz -98% False positive rate ≤5% 0 / 20 times -100%

[0301] Experimental conclusion: In the oscillating drift fault detection experiment, the system can effectively distinguish between "continuous oscillation" and "instantaneous interference" through the baseline rise rate, accurately identify oscillating drift faults, and the force control signal recovers stability after switching, verifying the effectiveness of the third fault judgment module.

[0302] The experimental results of four types of fault detection show that the dual-bridge redundant design and fault judgment module of the present invention can accurately detect open circuit, short circuit, unidirectional drift and oscillating drift faults, with fault response time of 5ms~28s (adapted according to fault type), which is better than the theoretical design target. After bridge switching, the force control fluctuation is stable at ±0.18N~±0.25N, with a false judgment rate of ≤5%, which fully meets the requirements of "force control reliability" and "miniaturization" in high-precision scenarios such as humanoid robot dexterous hands, medical equipment, and microelectronic press assembly. At the same time, the experimental data also verify the improvement effect of cross-beam type elastomer symmetrical patch design and integrated structure on measurement accuracy and spatial adaptability, proving the feasibility and superiority of the technical solution of the present invention.

Claims

1. A miniature servo electric cylinder, characterized in that, It includes a fault diagnosis module and push rods (9), hollow cup motor (4) and sensor flange (2) arranged from top to bottom. The output shaft of the hollow cup motor (4) and the push rod (9) are parallel to the vertical direction. The sensor flange (2) includes a horizontally arranged disc; Four first strain gauges (2d) are provided on the upper surface of the disk body, and the four first strain gauges (2d) are connected through resistors and amplifiers to form bridge circuit I; four second strain gauges (2e) are provided on the lower surface of the disk body, and the four second strain gauges (2e) are connected through resistors and amplifiers to form bridge circuit II; bridge circuit I and bridge circuit II are both Wheatstone bridges with full bridge connection; the first strain gauges (2d) and the second strain gauges (2e) are used to reflect the push and pull forces on the push rod (9); The fault diagnosis module includes a third fault diagnosis module, which is used to detect oscillating drift faults in bridge circuit I and bridge circuit II; the specific workflow is as follows: (1) Initialize parameters: Set t=0 and set the sampling time interval Δt; (2) Collect the instantaneous voltage value of bridge circuit I at time t. Instantaneous voltage value output by bridge circuit II ; (3) Calculate the predicted instantaneous voltage value at time t The residual of bridge I at time t The residual of bridge section II at time t The formula is as follows: ； ； ； In the formula, These are the preset fusion weight coefficients; The acquisition process is as follows: During the initialization phase of the micro servo electric cylinder, n instantaneous voltage values ​​output by bridge circuit I are continuously acquired, and their standard deviation is calculated. Given n values ​​≤ 200, synchronously collect n instantaneous voltage values ​​output by bridge circuit II and calculate their standard deviation. Substitute into the following formula to calculate: ； (4) Calculate the moving average of the voltage residual of bridge circuit I at time t. The moving average of the voltage residual of bridge circuit II at time t The formula is as follows: ； ； In the formula, This is a historical weighting factor, with a value range of 0.85-0.95; The previous time step at time t , The previous time step at time t When t=0, let =0, let =0; (5) Calculate the baseline rise rate of bridge section I at time t. The baseline rise rate of bridge section II at time t The formula is as follows: ； ； In the formula, Represents the current iteration number. Representing the During the next iteration The value, Representing the During the next iteration The value, Representing the During the next iteration The value of ; when t=0, let , ; (6) Fault diagnosis: judge and Whether they are true at the same time, The set threshold; If both conditions are met, it is determined that both Bridge I and Bridge II have experienced oscillatory drift, and the program terminates. If only If the condition is met, it is determined that bridge path I has experienced oscillatory drift, and the measurement work is switched to bridge path II, and the program terminates. If only If the condition is met, it is determined that bridge path II has experienced oscillatory drift, and the measurement work is switched to bridge path I, and the program terminates. If none of them are true, proceed to step (7); (7) Update sampling time: Let t = t + Δt, and return to step (2).

2. A miniature servo electric cylinder according to claim 1, characterized in that, The disc body of the sensor flange (2) consists of a ring and a cross-beam type elastic body located inside the ring; The cross-beam type elastomer is a one-piece molded part; The cross-beam elastic body consists of a load-bearing beam (2a), four deformable beams (2b) of equal length, and four hollow beams (2c) of equal length. The central axis of the load-bearing beam (2a) coincides with the central axis of the ring. The deformable beams (2b) and the hollow beams (2c) are arranged radially along the ring. One end of each of the four deformable beams (2b) is connected to the load-bearing beam (2a), and the other end is connected to one end of each of the four hollow beams (2c). The other end of each of the four hollow beams (2c) is connected to the inner wall of the ring. Four first strain gauges (2d) are respectively attached to the upper surface of four deformable beams (2b) and are evenly distributed in a 90° ring around the stressed beam (2a); Four second strain gauges (2e) are respectively attached to the lower surface of the four deformable beams (2b) and are symmetrically distributed with the four first strain gauges (2d).

3. A miniature servo electric cylinder according to claim 2, characterized in that, The upper surface of the load-bearing beam (2a) is a plane, and a threaded rod is fixed on the lower surface of the load-bearing beam (2a).

4. A miniature servo electric cylinder according to claim 1, characterized in that, It also includes a gearbox (6) and a trapezoidal screw (7); the sensor flange (2), the hollow cup motor (4), the gearbox (6), and the trapezoidal screw (7) are arranged from bottom to top; The gearbox (6) is equipped with an internal gear ring, a planetary gear (6a), a sun gear (6b) and a planetary gear carrier (6c). The sun gear (6b) meshes with the planetary gear (6a), the planetary gear (6a) meshes with the internal gear ring, and the planetary gear carrier (6c) connects the planetary gear (6a) and the sun gear (6b) in series. The trapezoidal screw (7) includes a flange portion, a stepped optical shaft portion and a threaded portion arranged from bottom to top; The output shaft of the hollow cup motor (4) is coaxial with the sun gear (6b) and the two are fixedly connected; the flange part of the trapezoidal screw (7) is connected to the planetary gear (6a); the push rod (9) is a cylindrical structure, the push rod (9) is sleeved on the threaded part of the trapezoidal screw (7) and threadedly connected to it; the lower end of the push rod (9) is equipped with an anti-rotation slider (9a).

5. A miniature servo electric cylinder according to claim 4, characterized in that, The anti-rotation slider (9a) has slots on both sides for embedded mounting of the first position sensor (9b) and the second position sensor (9c).

6. A miniature servo electric cylinder according to claim 4, characterized in that, The stepped optical shaft portion of the trapezoidal screw (7) is fitted with a thrust bearing (8), and the inner ring of the thrust bearing (8) is interference-fitted with the stepped optical shaft portion of the trapezoidal screw (7).

7. A miniature servo electric cylinder according to claim 6, characterized in that, It also includes a third end cap (11), a second housing (10), a first housing (5), a second end cap (3), and a first end cap (1); The third end cap (11), the second housing (10), the gearbox (6), the first housing (5), the second end cap (3), the sensor flange (2), and the first end cap (1) are connected sequentially from top to bottom; The first end cap (1) is fixedly connected to the ring of the sensor flange (2) by bolts; the second end cap (3) is connected to the first housing (5) by bolts; the third end cap (11) constrains the stroke of the push rod (9) by a limiting structure; The trapezoidal screw (7), the thrust bearing (8) and the push rod (9) are all located inside the second housing (10). The interior of the second housing (10) is provided with a circular groove and a square slide. The outer ring of the thrust bearing (8) is embedded in the circular groove. The push rod (9) is slidably connected to the square slide. The anti-rotation slider (9a) is inserted into the square slide. The main body of the hollow cup motor (4) is installed inside the first housing (5).

Images