Automatic hardware polishing method and system based on force feedback control

Abstract

The invention discloses an automatic hardware polishing method based on force feedback control, which comprises the following steps: mounting a topological quantum superconducting composite force sensor at the joint of a mechanical arm and a polishing tool to obtain an original quantum state signal; quantum entanglement is introduced to enhance transmission of original quantum state signals; training the quantum neural network; and multi-degree-of-freedom polishing is realized based on the quantum neural network. Transmission of original quantum state signals is enhanced through quantum entanglement, and the situation that the measurement precision is affected by sensor local de-coherence is avoided. Specifically, an entangled photon pair is generated by using an external quantum light source, and phase encoding is performed on an original force signal through an electro-optical modulator, so that the original force signal keeps entangled with standby photons, and the signal stability and the anti-noise capability are improved.

Description

Technical Field

[0001] The present invention belongs to the field of hardware polishing, and more specifically, relates to an automated hardware polishing method and system based on force feedback control. Background Art

[0002] In the field of high-end manufacturing, such as aerospace, precision optics, semiconductor manufacturing and other industries, extremely high requirements are placed on the surface quality of metal workpieces. Taking the blades of aero-engines as an example, the surface polishing quality directly affects the aerodynamic performance and service life of the engine. Minute defects (such as scratches, cracks, and micro-roughness) on the blade surface can cause airflow disturbances, reduce fuel efficiency, and may trigger fatigue damage, thus affecting the reliability of the entire engine. To meet the stringent surface quality requirements, blade polishing requires precise control of the normal force, tangential force, and torque force to ensure uniform and stable high-precision surface treatment without damaging the blade material.

[0003] Currently, the automated polishing of high-end precision hardware mainly relies on the mode of robotic arm + force sensor + control algorithm, among which force feedback control technology is the key to determining the polishing quality. Traditional force sensors mainly include resistance strain gauges, piezoelectric sensors, electromagnetic sensors, etc. These sensors are used to measure the force between the polishing tool and the workpiece, and adjust the polishing parameters through the control system to adapt to different surface topographies and material properties. In addition, some high-end equipment has introduced a fusion control method of vision detection + force feedback, which optimizes the polishing path and force distribution by optically measuring the surface roughness and combining the force sensor signals to improve the processing accuracy.

[0004] In terms of control algorithms, the current mainstream methods include PID control, adaptive force control, deep learning force control optimization, etc. These methods can reduce the force fluctuations during the polishing process to a certain extent and improve the processing consistency. However, due to the limited measurement accuracy and anti-interference ability of traditional force sensors, even in combination with advanced control algorithms, it is still difficult to solve the instability problem of the force feedback signal during the polishing process.

[0005] Although the existing force feedback control methods have made certain progress in industrial applications, there are still technical bottlenecks such as limited precision, serious environmental interference, and insufficient control stability in the field of high-end precision hardware polishing, which are mainly manifested as follows:

[0006] Limited accuracy of traditional force sensors: During the polishing process, traditional force sensors (such as strain gauges or piezoelectric sensors) are vulnerable to environmental noise, leading to increased measurement errors and difficulty in stably obtaining high-precision force feedback data. For example, in high-temperature, high-humidity, or strong electromagnetic field environments, the sensitivity of the sensor will drift, resulting in a decrease in the reliability of the feedback signal. In addition, the resolution of traditional sensors is usually at the micro-newton level, while high-end polishing processes require higher detection accuracy, and traditional sensors are difficult to meet the needs of ultra-precision machining.

[0007] Dispersive decoherence leads to reduced signal stability: In a high-precision force control system, the stability and consistency of the polishing force signal are crucial. However, due to the susceptibility of the sensor signal to environmental interference and the dynamic characteristics of the force signal itself, it is difficult to maintain the synchronization of the phase relationships of different force feedback signals in the system. This phenomenon is called "dispersive decoherence", that is, the force signal exhibits phase drift at different time points, making it difficult for the system to accurately predict the distribution of the polishing force. Due to the complex interference factors in the polishing environment, the signal decoherence problem is further exacerbated, resulting in a decrease in the robustness of the force feedback control system and affecting the final polishing quality. Summary of the Invention

[0008] To address the deficiencies in the prior art, the purpose of the present invention is to solve the above-mentioned defects and further propose a method and system for automatic polishing of hardware parts based on force feedback control.

[0009] The present invention adopts the following technical solutions.

[0010] The first aspect of the present invention discloses a method for automatic polishing of hardware parts based on force feedback control, including steps 1 to 4;

[0011] Step 1, install a topological quantum superconducting composite force sensor at the connection between the robotic arm and the polishing tool to obtain the original quantum state signal;

[0012] Step 2, introduce quantum entanglement to enhance the transmission of the original quantum state signal;

[0013] Step 3, train a quantum neural network;

[0014] Step 4, achieve multi-degree-of-freedom polishing based on the quantum neural network.

[0015] Furthermore, the forces detected by the topological quantum superconducting composite force sensor include: normal force, tangential force, and torque force, corresponding to the control parameters of pressure value, speed value, and angle value for multi-degree-of-freedom polishing respectively.

[0016] Furthermore, step 2 specifically includes steps 2.1 to 2.2;

[0017] Step 2.1, set an external quantum light source;

[0018] Step 2.2: Based on the electro-optic modulator, output an entangled state force signal to enhance the original quantum state signal.

[0019] Further, Step 3 specifically includes Step 3.1 to Step 3.2;

[0020] Step 3.1: Preprocess the entangled state force signal to obtain a quantum superposition state;

[0021] Step 3.2: Based on the trained quantum neural network, obtain the force probability distribution.

[0022] Further, Step 3.2 also includes Step S1 to Step S3;

[0023] Step S1: Install a cylindrical electromagnetic shielding cover around the topological quantum superconducting composite force sensor;

[0024] Step S2: Based on the magnetic field strength, predict the decoherence coefficient;

[0025] Step S3: Based on the decoherence coefficient and the main interference component in the polishing environment, obtain the adjusted force probability distribution.

[0026] Further, Step 4 specifically includes Step 4.1 to Step 4.3;

[0027] Step 4.1: Based on the current real-time entangled state force signal, decode the output force value of the quantum neural network;

[0028] Step 4.2: Based on the output force value, calculate the adjustment amount of the polishing parameters;

[0029] Step 4.3: Update the polishing parameters to achieve multi-degree-of-freedom polishing.

[0030] The second aspect of the present invention discloses a hardware part automatic polishing system based on force feedback control for executing the method described in the first aspect. The system includes: a robotic arm, a polishing tool, a topological quantum superconducting composite force sensor, a control system, a workpiece clamping device, an external quantum light source; an electro-optic modulator and a quantum neural network module;

[0031] The topological quantum superconducting composite force sensor is installed at the connection between the robotic arm and the polishing tool for obtaining the original quantum state signal;

[0032] The external quantum light source is used to generate entangled photon pairs, and based on the entangled photon pairs, the electro-optic modulator encodes the original quantum state signal into an entangled state force signal, thereby enhancing the transmission of the original quantum state signal;

[0033] The quantum neural network module trains the quantum neural network;

[0034] The control system is based on a quantum neural network and manipulates the robotic arm to achieve multi-degree-of-freedom polishing;

[0035] The workpiece clamping device is used to fix the hardware.

[0036] A third aspect of the present invention discloses a terminal, including a processor and a storage medium; characterized in that:

[0037] The storage medium is used to store instructions;

[0038] The processor is used to operate according to the instructions to execute the method described in the first aspect.

[0039] A fourth aspect of the present invention discloses a computer-readable storage medium, on which a computer program is stored, characterized in that the program, when executed by a processor, implements the method described in the first aspect.

[0040] The beneficial effects of the present invention are that, compared with the prior art, the present invention has the following advantages:

[0041] The present invention enhances the transmission of the original quantum state signal through quantum entanglement, avoiding the influence of sensor local decoherence on the measurement accuracy. Specifically, an external quantum light source is used to generate entangled photon pairs, and the original force signal is phase-encoded by an electro-optic modulator to make it remain entangled with the spare photons, thereby improving the signal stability and anti-noise ability.

[0042] After entering the quantum neural network, the entangled state force signal is converted into a voltage signal by a photodetector and normalized into four qubits of normal force, tangential force, torque force, and decoherence prediction. Among them, the first three correspond to the three key degrees of freedom of polishing force control, and the introduction of the decoherence interference measurement qubit is crucial for compensating the distortion of the force signal caused by environmental interference.

[0043] Since the decoherence effect will destroy the quantum superposition state and cause the force signal to deviate from the true value, it cannot be corrected in real time by relying solely on traditional anti-noise methods. The present invention stores the environmental noise information in qubit q[3], enabling the quantum neural network to dynamically adjust the force probability distribution during the learning process, predicting and compensating for the decoherence effect, thereby ensuring high-precision force decoding. This method improves the stability of the polishing force feedback, reduces measurement errors, enables the system to accurately control the polishing force in a complex industrial environment, and improves the processing quality. Description of the Drawings

[0044] Figure 1 It is a flowchart of an automatic polishing method for hardware based on force feedback control. Detailed Embodiments

[0045] The present application will be further described below in conjunction with the accompanying drawings. The following embodiments are only used to more clearly illustrate the technical solutions of the present invention and cannot be used to limit the protection scope of the present application.

[0046] In the prior art, the automated polishing system for hardware parts mainly includes the following components:

[0047] (1) Robot arm: Clamp the polishing tool or the hardware part, perform the polishing action, achieve multi-directional movement, and be driven by a motor and a transmission mechanism to provide precise movement with multiple degrees of freedom.

[0048] (2) Polishing tool: Such as a grinding wheel or a polishing wheel, directly polish the surface of the hardware part, remove burrs and defects, and use high-speed rotation or reciprocating motion to cooperate with the polishing agent to achieve surface treatment.

[0049] (3) Force sensor: Real-time detect the force acting on the hardware part during the polishing process, provide feedback data, and convert the mechanical force into an electrical signal through a strain gauge or piezoelectric material.

[0050] (4) Control system: Receive the signal from the force sensor, control the movement of the robot arm and the force of the polishing tool, achieve precise polishing, and form a force feedback closed-loop control based on algorithms such as PID control or fuzzy control.

[0051] (5) Workpiece clamping device: Fix the hardware part to ensure the stable position of the workpiece during the polishing process, and fix the workpiece through a clamping mechanism or vacuum adsorption technology.

[0052] In the embodiments of the present invention, the main types of hardware parts are titanium alloys (for example: Ti-6Al-4V). The blade lengths of these titanium alloys are usually 50 - 150 mm, and the thickness is 1 - 5 mm. Since the surface is usually a hyperboloid design, its curvature is relatively complex. Due to its high strength, corrosion resistance, and lightweight characteristics, it is widely used in the aerospace field (for example: turbine blades or compressor blades). However, its hardness and processing difficulty also require extremely high force control accuracy during the polishing process to avoid over-cutting or under-polishing.

[0053] The present invention discloses an automated polishing method for hardware parts based on force feedback control, as Figure 1 shown, including Step 1 to Step 4.

[0054] Step 1, install the topological quantum superconducting composite force sensor at the connection between the robot arm and the polishing tool to obtain the original quantum state signal.

[0055] In an embodiment of the present invention, a topological quantum superconducting composite force sensor (hereinafter referred to as a force sensor) mainly selects a Bi2Te3 film (thickness is about 70nm), which is prepared on a silicon substrate by chemical vapor deposition and cut into 5nm*5nm squares as a topological quantum material layer. Bi2Te3 is a typical three-dimensional topological insulator, and its unique surface state can produce a significant electrical signal response under a small force change. The topological quantum material layer is surrounded by a miniature superconducting coil, which is usually prepared from NbTi material, with a diameter of 1mm, a number of turns set to 50, and fixed by a low-temperature evaporation process. When in use, a constant current is passed to generate a local magnetic field B (approximately 0.1T), thereby enhancing the edge state against thermal noise and electromagnetic interference.

[0056] The significance of using topological quantum material force sensors lies in: (1) Ultra-high sensitivity requirements: The complex curved surface and micron-level tolerance of turbine blades require that the force control system be able to sense and adjust tiny force deviations in real time to avoid surface quality defects. (2) Stability in complex environments: During the polishing process, the contact between the grinding tool and the titanium alloy will generate local thermal noise (temperature rise of 5-20°C) and electromagnetic fluctuations (such as interference from motor drives). Traditional sensor signals are easily distorted, while the topological protection characteristics of topological quantum states can theoretically resist these decoherence effects. (3) Real-time feedback requirements: The polishing robot needs to adjust the force output every millisecond to adapt to the change in blade curvature. The high response speed of the topological quantum sensor (thanks to quantum state conduction) can support this requirement.

[0057] In some embodiments, the forces that the force sensor needs to detect mainly include: normal force, that is, the vertical contact force between the polishing tool and the surface of the hardware, which determines the polishing depth and quality; tangential force, that is, the friction force when the polishing tool moves along the surface of the hardware, which affects the polishing efficiency and surface finish; torque force: the rotational force generated when the polishing tool rotates, which affects the stability and service life of the tool. Normal force, tangential force and torque force correspond to the control parameters of multi-degree-of-freedom polishing, namely pressure value, speed value and angle value.

[0058] It is understandable that the essence of the force sensor is to convert the normal force, tangential force and torque force into force-electrical signals, that is, the output is the original quantum state signal. The force sensor is precisely installed at the connection between the robot arm and the polishing tool. This position can directly sense the dynamic contact force between the tool and the workpiece. To ensure the accuracy of the measurement, high-strength bolts are used to fix the sensor during installation, and the coaxiality with the robot arm is adjusted through precision calibration equipment (error less than 0.1 microns). In addition, mechanical vibration or looseness must be avoided during installation to ensure the stability of the sensor output signal.

[0059] Step 2: Enhance the transmission of the original quantum state signal based on quantum entanglement.

[0060] Step 2 utilizes the non-locality of quantum entanglement to avoid the direct influence of local decoherence of the sensor on the signal output. Specifically, it includes Step 2.1 to Step 2.2.

[0061] Step 2.1, set up an external quantum light source.

[0062] In some embodiments, a non-linear crystal (such as β-barium borate BBO, size 10mm × 5mm × 1mm) can be used as the external quantum light source, which is set on a fixed bracket beside the workpiece clamping device. The crystal is excited by a pump laser (wavelength 405nm, power 50mW) to generate entangled photon pairs (wavelength 810nm). The distance between the light source and the sensor is controlled within 10 cm to ensure the signal coupling efficiency.

[0063] Step 2.2, based on the electro-optic modulator, output an entangled state force signal to enhance the original quantum state signal.

[0064] Specifically, the original quantum state signal is input into the electro-optic modulator. The original quantum state signal drives the electro-optic modulator to modulate the photon phase, generating a phase difference. The modulated photon remains entangled with the spare photon to form an encoded entangled state force signal.

[0065] It can be understood that "encoding" means modulating the original force signal onto the optical parameters of the entangled photon pair through the electro-optic modulator, specifically modulating the phase difference of the photon, so as to generate an entangled state force signal. Due to the non-locality of the entangled state, the signal is not easily affected by environmental noise (such as electromagnetic interference) during transmission, thereby improving the anti-noise ability and stability.

[0066] Step 3, train the quantum neural network.

[0067] The significance of Step 3 is to convert the optical signal, that is, the entangled state force signal, into an executable instruction, so as to intelligently compensate for the decoherence effect. Specifically, it includes Step 3.1 to Step 3.2.

[0068] Step 3.1, preprocess the entangled state force signal to obtain a quantum superposition state.

[0069] Specifically, the entangled state force signal passes through a photodetector to obtain a voltage signal. After normalizing the voltage signal, 4 quantum bits are used, corresponding to q[0]: normal force; q[1]: tangential force; q[2]: torque force; q[3]: decoherence prediction respectively. And it is encoded into a 4-bit quantum superposition state Qe through the rotation gate RY.

[0070] That is to say, the quantum superposition state includes qubits in four dimensions of normal force, tangential force, torque force, and decoherence prediction. The present invention stores environmental noise information through qubit q[3], enabling the quantum neural network to dynamically adjust the force probability distribution during the learning process, predict and compensate for the decoherence effect, thereby ensuring high-precision force decoding.

[0071] The rotation gate is a single-qubit gate in quantum computing. For example, RY is the state of the qubit rotating by an angle θ around the Y-axis of the Bloch Sphere. The formula for θ is:

[0072] θ = π × Vn

[0073] Since the range of the normalized voltage signal Vn is from 0 to 1, the range of θ is [0, π).

[0074] |ψ_i> = cos(θ / 2)|0> + sin(θ / 2)|1>

[0075] Qe = [|ψ_1>, |ψ_2>, |ψ_3>, |ψ_4>]

[0076] Where |0> and |1> are Dirac notations in quantum mechanics, representing the "0" state and the "1" state respectively. |ψ_i> represents the state of qubit q[i], where i = 1, 2, 3, 4. It can be understood that since the θ applied to each qubit is the same, the superposition states of the four qubits are exactly the same before learning through the quantum neural network.

[0077] Step 3.2, based on the trained quantum neural network, obtain the force probability distribution.

[0078] Build a quantum neural network on the Qiskit platform, including an input layer, namely the quantum superposition state Qe; a hidden layer, namely four parameterized quantum gates, including: two CNOT gates and two RZ gates; an output layer, namely the Z-basis measurement. The hidden layer processes the quantum superposition state through a variational quantum circuit, and uses the superposition state to parallelly decode multi-dimensional force data. The training data can be collected based on the polishing experiment in the static state to output the force probability distribution P(F) = [p[0], p[1], p[2]], where p[i] represents the probability of measuring q[i] respectively, and i = 1, 2, 3.

[0079] It can be understood that RZ is the state of the qubit rotating by an angle θ around the Z-axis of the Bloch Sphere. The CNOT gate is a two-qubit gate, called the controlled-NOT gate. It determines whether to flip the state of the qubit according to the state of the control qubit ("0" state and "1" state). The Z-basis measurement means that in quantum computing, the qubit is measured along the basis state of the Z-axis, causing the measurement result to collapse into a specific binary value.

[0080] The training method is as follows:

[0081] Use the classical optimizer COBYLA and set the parameters: iterate 200 times with a learning rate of 0.01; minimize the loss function L = SUM((Fm - Fr) 2 ), where Fm is the output force value of the quantum neural network, representing the intermediate estimated value of the quantum neural network, and Fr is the calibrated force value measured experimentally in the stationary state.

[0082] Understandably, the trained quantum neural network is used to decode the output force value of the quantum neural network based on the current real-time entangled state force signal.

[0083] In some embodiments, EMC interference compensation can be utilized to compensate the force probability distribution, thereby obtaining a more accurate force probability distribution. Step 3.2 further includes steps S1 to S3.

[0084] Step S1: Install a cylindrical electromagnetic shielding cover around the topological quantum superconducting composite force sensor.

[0085] Among them, the material of the electromagnetic shielding cover is high-permeability μ-metal, with a diameter of 5 cm and a thickness of 2 mm, and an adjustable Faraday cage is embedded, with a mesh aperture of 0.5 mm. The shielding cover is driven by an external power supply to generate a shielding frequency in the range of 10 kHz - 1 MHz to block the electromagnetic interference of the polishing motor.

[0086] Understandably, the significance of step S1 is to more effectively protect the entangled state force signal from external electromagnetic noise, and the above-mentioned electromagnetic noise mainly comes from the high-speed rotation of the polishing tool.

[0087] Step S2: Predict the decoherence coefficient based on the magnetic field strength. The decoherence coefficient Dp is shown as follows:

[0088] Dp = p[3] × Kr × (1 - Ba / Bm)

[0089] Where Kr is the decoherence rate constant, which can be calibrated through experiments, Ba and Bm are the average magnetic field strength and the maximum magnetic field strength of the electromagnetic shielding cover respectively, and p[3] is the probability of measuring q[3].

[0090] In some embodiments, q[3] can be the decoherence coefficient Dp.

[0091] Step S3: Obtain the adjusted force probability distribution based on the decoherence coefficient and the main interference component in the polishing environment.

[0092] In some embodiments, an electromagnetic interference sensor can be used to measure the polishing environmental interference power Pe, and the main interference component ΔP is extracted through a band-pass filter (center frequency set to 50 kHz and bandwidth of 10 kHz) to obtain the adjusted force probability distribution P’(F). P’(F) is shown as follows:

[0093] P’(F) = P(F) × (1 - Dp-emc × ΔP)

[0094] Where emc is the EMC compensation coefficient, which can be determined multiple times using an electromagnetic interference sensor under the condition of no polishing operation and taking the average value.

[0095] The relevant source code for step 3 can be shown as follows:

[0096] from qiskit import QuantumCircuit,Aer,execute

[0097] from qiskit.circuit import ParameterVector

[0098] import numpy as np

[0099] # Define the number of qubits

[0100] num_qubits = 4 # Corresponding to q[0] to q[3]

[0101] # Create a quantum circuit

[0102] qc = QuantumCircuit(num_qubits)

[0103] # Define the trainable parameter theta

[0104] theta = ParameterVector('theta', length = 4) # Corresponding to the parameters θ1 to θ4 of the RZ gate

[0105] # Assume that the input quantum superposition state has been encoded through the RY gate (such as encoding the force signal Vn)

[0106] # Here, construct the hidden layer and output layer of the QNN

[0107] qc.ry(np.pi / 4, range(num_qubits)) # Example: Initial superposition state encoding

[0108] # Hidden layer: Add RZ gates and CNOT gates

[0109] qc.rz(theta[0],0) #RZ(θ1) acts on q[0]

[0110] qc.rz(theta[1],1) #RZ(θ2) acts on q[1]

[0111] qc.cx(0,1) # CNOT gate, controlling q[0], targeting q[1]

[0112] qc.cx(1,2) # CNOT gate, controlling q[1], targeting q[2]

[0113] # Output layer: Z-basis measurement

[0114] qc.measure_all() # Measure all qubits

[0115] # Bind parameters

[0116] theta_values = [0.5, 0.3, 0.2, 0.1]

[0117] bound_qc = qc.bind_parameters({theta: theta_values})

[0118] # Execute the circuit and measure

[0119] simulator = Aer.get_backend('qasm_simulator')

[0120] shots = 1024

[0121] result = execute(bound_qc, simulator, shots = shots).result()

[0122] counts = result.get_counts()

[0123] # Calculate the force probability distribution P(F)

[0124] p0 = sum(counts.get(state, 0) for state in counts if state[-1] == '0') / shots # Probability that q[0] == 0

[0125] p1 = sum(counts.get(state, 0) for state in counts if state[-2] == '0') / shots # Probability that q[1] == 0

[0126] p2 = sum(counts.get(state, 0) for state in counts if state[-3] == '0') / shots # q[2]=0 probability

[0127] P_F = [p0, p1, p2] # Force probability distribution

[0128] print("Force probability distribution P(F):", P_F)

[0129] Step 4: Implement multi-degree-of-freedom polishing based on a quantum neural network.

[0130] Step 4.1, Based on the current real-time entangled state force signal, decode the output force value of the quantum neural network.

[0131] It can be understood that the output force value F = [Fn, Ft, Fo], where Fn, Ft, and Fo represent the normal force, tangential force, and torque force respectively.

[0132] Step 4.2, Based on the output force value, calculate the adjustment amount of the polishing parameters.

[0133] In the embodiment of the present invention, the polishing parameters mainly include Pc, Vc, and Ac, which represent the pressure value, speed value, and angle value respectively.

[0134] The adjustment amounts ΔP, ΔV, and ΔA of the polishing parameters are shown in the following formula:

[0135] ΔP = kp × (Fn’ - Fn)

[0136] ΔV = kv × (Ft’ - Ft)

[0137] ΔA = ka × (Fo’ - Fo)

[0138] Among them, kp, kv, and ka are the pressure control gain, speed control gain, and angle control gain respectively, which are obtained by experimental calibration; Fn’, Ft’, and Fo’ are the target normal force, target tangential force, and target torque force respectively, which are preset according to process requirements.

[0139] Step 4.3, Update the polishing parameters to achieve multi-degree-of-freedom polishing.

[0140] It can be understood that the new polishing parameters are shown in the following formula:

[0141] P[k + 1] = P[k] + ΔP

[0142] V[k + 1] = V[k] + ΔV

[0143] A[k + 1] = A[k] + ΔA

[0144] Wherein, k = 0, 1, 2,... represents the frequency of polishing control. Therefore, P[0], V[0], and A[0] respectively represent the polishing parameters at the initial stage, and P[k], V[k], and A[k] respectively represent the polishing parameters at the k-th frequency.

[0145] Correspondingly, the present invention also discloses an automated polishing system for hardware based on force feedback control, including: a robotic arm, a polishing tool, a topological quantum superconducting composite force sensor, a control system, a workpiece clamping device, and an external quantum light source; an electro-optic modulator and a quantum neural network module;

[0146] The topological quantum superconducting composite force sensor is installed at the connection between the robotic arm and the polishing tool for obtaining the original quantum state signal;

[0147] The external quantum light source is used to generate entangled photon pairs, and based on the entangled photon pairs, the electro-optic modulator encodes the original quantum state signal into an entangled state force signal, thereby enhancing the transmission of the original quantum state signal;

[0148] The quantum neural network module trains the quantum neural network;

[0149] The control system, based on the quantum neural network, controls the robotic arm to achieve multi-degree-of-freedom polishing;

[0150] The workpiece clamping device is used to fix the hardware.

[0151] Finally, it should be noted that the meanings of "quantum superposition state", "dimension", "decoherence prediction", etc. involved in the present invention are not consistent with the meanings of the quantum superposition state in true quantum mechanics. In fact, at least in the content of the entire step 3, quantum mechanics is not involved; that is to say, the meanings of "quantum superposition state", "dimension", "decoherence prediction", etc. are only used to metaphorically introduce the quantum neural network (substantially a multi-dimensional state analysis method based on qubits) to make it more convenient to describe.

[0152] The applicant of the present invention has made a detailed description and explanation of the implementation examples of the present invention in combination with the accompanying drawings of the specification. However, those skilled in the art should understand that the above implementation examples are only the preferred implementation schemes of the present invention, and the detailed description is only to help readers better understand the spirit of the present invention, rather than a limitation on the protection scope of the present invention. On the contrary, any improvement or modification made based on the spirit of the present invention should fall within the protection scope of the present invention.

Claims

1. An automated polishing method for hardware parts based on force feedback control, characterized in that It includes steps 1 to 4; Step 1: Install a topological quantum superconducting composite force sensor at the connection between the robotic arm and the polishing tool to obtain the original quantum state signal; Step 2: Introduce quantum entanglement to enhance the transmission of the original quantum state signal; Step 3: Train a quantum neural network; Step 4: Achieve multi-degree-of-freedom polishing based on the quantum neural network.

2. The automated polishing method for hardware based on force feedback control according to claim 1, characterized in that The forces detected by the topological quantum superconducting composite force sensor include: normal force, tangential force, and torque force, which respectively correspond to the control parameters of pressure value, speed value, and angle value for multi-degree-of-freedom polishing.

3. A method for automatically polishing hardware based on force feedback control according to claim 1, characterized in that, Step 2 specifically includes steps 2.1 to 2.2; Step 2.1: Set up an external quantum light source; Step 2.2: Based on an electro-optic modulator, output an entangled state force signal to enhance the original quantum state signal.

4. A method for automatically polishing hardware based on force feedback control according to claim 1, characterized in that, Step 3 specifically includes steps 3.1 to 3.2; Step 3.1: Preprocess the entangled state force signal to obtain a quantum superposition state; Step 3.2: Based on the trained quantum neural network, obtain the force probability distribution.

5. A method for automatically polishing hardware based on force feedback control according to claim 4, characterized in that, Step 3.2 further includes steps S1 to S3; Step S1: Install a cylindrical electromagnetic shielding cover around the topological quantum superconducting composite force sensor; Step S2: Predict the decoherence coefficient based on the magnetic field strength; Step S3: Based on the decoherence coefficient and the main interference components in the polishing environment, obtain the adjusted force probability distribution.

6. A method for automatic polishing of hardware based on force feedback control according to claim 1, characterized in that, Step 4 specifically includes steps 4.1 to 4.3; Step 4.1: Based on the current real-time entangled state force signal, decode the output force value of the quantum neural network; Step 4.2: Based on the output force value, calculate the adjustment amount of the polishing parameters; Step 4.3: Update the polishing parameters to achieve multi-degree-of-freedom polishing.

7. An automated polishing system for hardware based on force feedback control, which is used to execute the method described in any one of claims 1-6, and is characterized in that The system includes: a robotic arm, a polishing tool, a topological quantum superconducting composite force sensor, a control system, a workpiece clamping device, and an external quantum light source; an electro-optic modulator and a quantum neural network module; The topological quantum superconducting composite force sensor is installed at the connection between the robotic arm and the polishing tool for obtaining the original quantum state signal; The external quantum light source is used to generate entangled photon pairs, and based on the entangled photon pairs, the electro-optic modulator encodes the original quantum state signal into an entangled state force signal, thereby enhancing the transmission of the original quantum state signal; The quantum neural network module trains the quantum neural network; The control system manipulates the robotic arm to achieve multi-degree-of-freedom polishing based on the quantum neural network; The workpiece clamping device is used to fix the hardware.

8. A terminal, including a processor and a storage medium; characterized in that: The storage medium is used to store instructions; The processor is used to operate according to the instructions to execute the steps of the method according to any one of claims 1-6.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the program is executed by the processor, it implements the steps of the method according to any one of claims 1-6.

Images