An intelligent speed control system for a hydraulic equipment for disassembling automobiles

Abstract

The application relates to the technical field of engineering machinery control, and discloses an intelligent speed regulation control system for automobile disassembly hydraulic equipment, which comprises a pressing arm assembly, a shearing arm assembly and an edge controller. A proportional overflow valve and an acceleration sensor are arranged in a pressing arm oil return circuit, and an electro-hydraulic servo valve and a pressure sensor are arranged in a shearing oil inlet circuit. The controller superimposes a high-frequency excitation feature in an electro-hydraulic servo valve instruction, collects shearing pressure and pressing arm vibration signals, and calculates a workpiece equivalent stiffness index through frequency domain analysis. Based on the index, the controller outputs an adaptive flow instruction negatively correlated with the stiffness to the electro-hydraulic servo valve and outputs a dynamic back pressure instruction introducing virtual damping to the proportional overflow valve. The application realizes adaptive matching of shearing speed and pressing arm damping characteristics by online identification of workpiece stiffness, solves the problems of low operation efficiency and impact damage under different working conditions, and improves the system control precision and safety.

Description

Technical Field

[0001] This invention relates to the field of engineering machinery control technology, specifically to an intelligent speed control system for hydraulic equipment used in automobile dismantling. Background Technology

[0002] The objects targeted in end-of-life vehicle dismantling operations include chassis beams, engine blocks, body panels, interior components, and other parts. The material strength and structural rigidity vary among these different components. Existing hydraulic dismantling equipment typically employs constant displacement control, constant power control, or relies on manual adjustment of valve openings by operators. This control mode struggles to adapt to the changing load characteristics during the dismantling process in real time.

[0003] In actual operation, when the equipment switches from shearing low-stiffness components to shearing high-strength components, if the system maintains a high flow rate, the hydraulic system pressure can easily rise rapidly to the overflow limit, causing the engine to slow down or stall. Conversely, when handling low-stiffness components, if common control parameters are used, the speed advantage of the hydraulic system cannot be fully utilized, leading to reduced work efficiency. Furthermore, during the shearing of high-stiffness components, the workpiece releases elastic potential energy upon fracture. Because the existing equipment's pressure arm mechanism lacks an active stiffness adjustment and damping absorption mechanism, this energy release is converted into hydraulic shock and mechanical vibration. Sustained impact loads can easily lead to failure of hydraulic cylinder seals and fatigue cracks in mechanical structural components, affecting the equipment's service life. Current technology lacks a control scheme that can identify the physical characteristics of the workpiece online during operation and adaptively adjust the flow rate and damping parameters accordingly. Summary of the Invention

[0004] To address the shortcomings of existing technologies, this invention provides an intelligent speed control system for automotive dismantling hydraulic equipment. This system solves the problems of low operating efficiency and severe mechanical impact damage caused by the inability of existing automotive dismantling hydraulic equipment to adaptively adjust hydraulic flow and back pressure damping parameters in real time when dealing with dismantling objects with different stiffness characteristics.

[0005] To achieve the above objectives, the present invention provides the following technical solution:

[0006] The first aspect of this invention provides an intelligent speed control system for automotive dismantling hydraulic equipment. This system includes a pressure arm assembly, a shearing arm assembly, a temperature sensor, and an edge controller. The pressure arm hydraulic cylinder's return oil circuit is equipped with a proportional relief valve, and an acceleration sensor is located at the end of the pressure arm. The shearing hydraulic cylinder's main oil inlet circuit is equipped with an electro-hydraulic servo valve, and a pressure sensor is located inside the cylinder. The edge controller connects to the aforementioned actuators and sensors via a communication bus.

[0007] The edge controller is configured to execute a cooperative control strategy: responding to external operation commands, it outputs valve port control commands superimposed with high-frequency excitation characteristics to the electro-hydraulic servo valve, driving the shearing hydraulic cylinder to generate hydraulic pulsation; it receives time-domain vibration signals from an accelerometer and pressure signals from a pressure sensor, and calculates the equivalent stiffness index of the workpiece; based on the equivalent stiffness index and time-domain vibration signals, it generates adaptive flow feedforward commands for the electro-hydraulic servo valve and dynamic back pressure setting commands for the proportional relief valve, respectively. The adaptive flow feedforward command is negatively correlated with the equivalent stiffness index, and the dynamic back pressure setting command is configured to absorb mechanical vibration energy.

[0008] The second aspect of this invention describes the specific processing logic of the above system when performing intelligent speed control.

[0009] In the excitation signal generation logic, the controller acquires the static bias signal required to maintain the current contact pressure. This signal value is greater than the dead zone initiation threshold of the electro-hydraulic servo valve. Based on this, a periodic fluctuation component is superimposed. The frequency of the fluctuation component is lower than the upper limit of the frequency response bandwidth of the electro-hydraulic servo valve and avoids the set of natural frequencies of the mechanical structure. Its amplitude is set to establish a dynamic pressure wave inside the shearing hydraulic cylinder without causing macroscopic displacement of the shearing arm assembly.

[0010] In the workpiece stiffness calculation logic, the controller performs windowing modulation and fast discrete Fourier transform on the time-domain vibration signal and pressure signal to generate frequency-domain power spectral density data. The input pressure amplitude and output acceleration amplitude at the detection frequency are extracted. Based on the input pressure amplitude, the effective bearing area of ​​the hydraulic cylinder, and friction loss force, the effective dynamic excitation force is calculated. The inertial response force of the pressure arm assembly is calculated based on the output acceleration amplitude. The ratio of the inertial response force to the effective dynamic excitation force is calculated to obtain the mechanical transmissibility, which is then normalized and mapped to an equivalent stiffness exponent.

[0011] Before performing stiffness calculations, the system verifies the effectiveness of physical coupling. The ratio of the total power of the main lobe signal in the neighborhood of the detection frequency to the average noise power density in the noise reference frequency bands on both sides is calculated as the frequency domain signal-to-noise ratio (SNR) indicator. If this indicator is greater than a preset confidence threshold, the physical coupling is considered effective; otherwise, it is considered invalid, and a default safe flow command is generated.

[0012] In the flow control command generation logic, the system parses the operation command to obtain the normalized flow demand command and calculates the flow correction gain coefficient using a nonlinear attenuation model based on a sigmoid function. The sigmoid function is configured to cause the flow correction gain coefficient to decrease rapidly after the equivalent stiffness exponent exceeds the inflection point stiffness value of the operating condition. The system also introduces pressure saturation correction logic to reduce the flow output when the actual shear pressure approaches the system overflow pressure. The final valve opening flow command is obtained by multiplying the rated flow rate, the flow demand command, the flow correction gain coefficient, and the pressure saturation correction coefficient.

[0013] In the back pressure control command generation logic, the system calculates the static reference pressure based on the equivalent stiffness exponent, which is positively correlated with the equivalent stiffness exponent. Simultaneously, the vibration velocity is obtained by integrating and high-pass filtering the time-domain vibration signal. This vibration velocity is then multiplied by a virtual damping coefficient to obtain the dynamic damping pressure compensation value. The static reference pressure and the dynamic damping pressure compensation value are then superimposed to serve as the basic control command for the proportional relief valve.

[0014] In addition, the system is equipped with fracture impact suppression logic, timing synchronization logic, and fault-oriented safety logic. The fracture impact suppression logic monitors the time-dependent component of the shear pressure signal and superimposes a pulse lockout pressure when workpiece fracture is detected. The timing synchronization logic estimates the system response lag time based on hydraulic oil temperature and system pressure, calculates the phase lead step, and preloads control parameters to align the control command activation time with the physical action time. The fault-oriented safety logic periodically verifies sensor signal characteristics and, upon detecting an anomaly, disables the adaptive control strategy, locking into manual speed limiting and safety pressure holding mode.

[0015] This invention provides an intelligent speed control system for hydraulic equipment used in automotive dismantling. It offers the following advantages:

[0016] 1. This invention achieves online quantitative identification of the equivalent stiffness of a workpiece by superimposing a high-frequency excitation signal that does not induce macroscopic displacement into the control command of an electro-hydraulic servo valve, and by using frequency domain analysis to calculate the transmissibility between input pressure and output vibration. Based on this stiffness index, the system constructs a negatively correlated flow rate mapping relationship. When processing high-stiffness workpieces such as chassis beams, the flow rate is automatically reduced to ensure stable shear force output; when processing low-stiffness workpieces such as thin plates, the flow rate is increased to accelerate the work cycle. This achieves adaptive working condition matching for different disassembly objects on the same equipment, improving overall work efficiency.

[0017] 2. This invention utilizes a proportional relief valve in the return oil circuit of the hydraulic cylinder of the pressure arm and an acceleration sensor at the end of the pressure arm to establish a dynamic back pressure control mechanism based on vibration velocity feedback. This mechanism generates a pressure compensation value by multiplying the vibration velocity by a virtual damping coefficient, enabling the hydraulic cylinder of the pressure arm to possess variable damping characteristics and actively absorb the mechanical vibration energy during the shearing process. Combined with fracture impact suppression logic based on pressure differential monitoring, the overflow threshold is rapidly increased at the moment of workpiece fracture, effectively suppressing the elastic rebound impact of the hydraulic cylinder and reducing fatigue damage to the mechanical structure.

[0018] 3. This invention integrates physical coupling effectiveness verification and control timing synchronization logic, improving the robustness and accuracy of the control system. By calculating the frequency domain signal-to-noise ratio to eliminate spurious signals caused by no-load or poor contact, the reliability of stiffness calculation is ensured. By collecting oil temperature and pressure data in real time to estimate the response lag time of the hydraulic system and performing phase lead compensation, the control delay caused by changes in oil viscosity and bulk modulus is eliminated, ensuring that the effective time of control commands is precisely aligned with the physical action time of the hydraulic system. Attached Figure Description

[0019] Figure 1 This is a schematic diagram of the algorithm flow of the present invention;

[0020] Figure 2 This is a schematic diagram of the system architecture of the present invention;

[0021] Figure 3 This is a comparison diagram of the vibration response power spectral density under different operating conditions of the present invention;

[0022] Figure 4 This is a comparison diagram of the multidimensional state response during the shearing operation process of the present invention;

[0023] Figure 5 This is a diagram showing the pressure impact response of the hydraulic cylinder of the pressure arm at the moment of workpiece fracture, as presented in this invention.

[0024] The components include: a pressure arm assembly 20, a shear arm assembly 30, an edge controller 100, a fieldbus 200, a workpiece 40, a pressure arm hydraulic cylinder 21, a proportional relief valve 22, an acceleration sensor 23, an electro-hydraulic servo valve 32, a shear hydraulic cylinder 31, an excitation generation module 101, a feature extraction module 102, an impedance identification module 103, a collaborative calculation module 104, and a pressure sensor 33. Detailed Implementation

[0025] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0026] Please see the appendix Figure 1 The present invention provides an intelligent speed control system for automotive dismantling hydraulic equipment, which is configured to perform clamping, fixing and shearing operations on unstructured metal loads.

[0027] The control system of the automotive dismantling hydraulic equipment includes a pressure arm assembly 20, a shear arm assembly 30, an edge controller 100, and a fieldbus 200. The pressure arm assembly 20 and the shear arm assembly 30 are mechanically connected to the same equipment base and are configured to contact different parts of the workpiece 40 during operation, thereby forming a physical and mechanical transmission link through the workpiece 40.

[0028] The pressure arm assembly 20 includes a pressure arm hydraulic cylinder 21, a proportional relief valve 22, and an acceleration sensor 23. The pressure arm hydraulic cylinder 21 drives the pressure arm assembly 20 to apply holding pressure to the workpiece 40. The proportional relief valve 22 is installed on the rodless chamber hydraulic circuit of the pressure arm hydraulic cylinder 21 and is electrically connected to the edge controller 100. It is used to adjust the return oil back pressure threshold of the pressure arm hydraulic cylinder 21 in real time according to received control commands. The acceleration sensor 23 is rigidly fixed to the end mechanical structure of the pressure arm assembly 20 and is used to collect vibration acceleration signals in three degrees of freedom directions at the end of the pressure arm assembly 20.

[0029] The shearing arm assembly 30 includes a shearing hydraulic cylinder 31, an electro-hydraulic servo valve 32, and a pressure sensor 33. The shearing hydraulic cylinder 31 drives the shearing arm assembly 30 to perform shearing actions on the workpiece 40. The electro-hydraulic servo valve 32 is installed in the main oil inlet of the shearing hydraulic cylinder 31 and is electrically connected to the edge controller 100. The electro-hydraulic servo valve 32 is configured not only to regulate the flow rate of hydraulic oil entering the shearing hydraulic cylinder 31 but also to superimpose micro-pulsation excitations on the output flow rate in response to high-frequency control signals. The pressure sensor 33 is installed at the oil inlet or rodless chamber of the shearing hydraulic cylinder 31 to collect real-time data on changes in the working pressure inside the shearing hydraulic cylinder 31.

[0030] The edge controller 100 is an embedded industrial control unit deployed locally on the device. The edge controller 100 establishes bidirectional communication connections with the proportional relief valve 22, the accelerometer 23, the electro-hydraulic servo valve 32, and the pressure sensor 33 via the fieldbus 200. The fieldbus 200 employs a communication protocol that supports deterministic low-latency transmission, ensuring that the time synchronization accuracy between sensor data acquisition and actuator command issuance is within the millisecond range.

[0031] The edge controller 100 is internally configured with a data processing logic unit, which includes an excitation generation module 101, a feature extraction module 102, an impedance identification module 103, and a collaborative computing module 104.

[0032] The excitation generation module 101 is used to generate a valve opening command superimposed with a preset frequency chatter signal, and send the command to the electro-hydraulic servo valve 32.

[0033] The feature extraction module 102 is used to receive the time-domain vibration signal fed back by the accelerometer 23 and perform fast Fourier transform processing on the signal to obtain frequency domain power spectral density data.

[0034] Impedance identification module 103 is used to calculate the transfer function between input excitation and output response, and to determine the validity of physical coupling connection based on the signal-to-noise ratio of frequency domain signal.

[0035] The collaborative calculation module 104 is used to look up a preset mapping table based on the impedance identification results, and generate flow feedforward commands for the electro-hydraulic servo valve 32 and dynamic back pressure setting commands for the proportional relief valve 22, respectively.

[0036] See attached document Figure 2 This invention provides an intelligent speed control method for automotive dismantling hydraulic equipment, comprising the following steps:

[0037] First, the edge controller 100 controls the pressure arm assembly 20 to preset the initial clamping pressure. The workpiece 40 is clamped and fixed, and the shearing arm assembly 30 is controlled to move towards the workpiece 40 until the shearing hydraulic cylinder 31 establishes a preset contact pressure. This creates a series mechanical transmission link between the pressure arm assembly 20, the workpiece 40, and the shear arm assembly 30 in physical space.

[0038] Subsequently, the edge controller 100 calls the excitation generation module 101 to generate a valve port control command containing high-frequency chatter characteristics, and sends the valve port control command to the electro-hydraulic servo valve 32 of the shear arm assembly 30 to drive the shearing hydraulic cylinder 31 to apply a small amplitude and a specific frequency to the workpiece 40. Hydraulic pulsation excitation;

[0039] Next, the edge controller 100 acquires the vibration response signal at the end of the pressure arm assembly 20 through the accelerometer 23, calls the feature extraction module 102 to perform a fast Fourier transform on the vibration response signal to generate frequency domain power spectral density data, and calls the impedance identification module 103 to calculate the excitation frequency. The ratio of the response amplitude at a given point to the ambient noise level;

[0040] When the ratio is greater than the preset signal-to-noise ratio threshold, the impedance identification module 103 determines that the physical coupling is effective and calculates the equivalent stiffness index of the workpiece 40. The collaborative computing module 104 calculates the equivalent stiffness index. An adaptive flow feedforward command is generated for the electro-hydraulic servo valve 32 and an active energy-absorbing back pressure command is generated for the proportional relief valve 22, wherein the active energy-absorbing back pressure command is configured to reduce the overflow opening threshold of the pressure arm hydraulic cylinder 21 under high-rigidity workpiece conditions.

[0041] Conversely, when the ratio is less than or equal to the preset signal-to-noise ratio threshold, the impedance identification module 103 determines that the physical coupling is invalid, and the collaborative calculation module 104 generates the default safe flow command and the maximum holding pressure command.

[0042] Following this, the edge controller 100 executes the timing handshake logic, first sending a back pressure setting command to the proportional relief valve 22 and executing it for a duration of [duration missing]. The forced delay waiting time is based on the hydraulic system's pressure-based response time setting;

[0043] Finally, after the timer ends, the edge controller 100 sends a flow feedforward command or a safe flow command to the electro-hydraulic servo valve 32 to drive the shearing hydraulic cylinder 31 to perform the shearing action. When the workpiece 40 breaks and generates impact energy, the pressure arm hydraulic cylinder 21, which is in a low back pressure setting state, absorbs the mechanical impact energy through overflow unloading.

[0044] The edge controller 100 controls the shearing hydraulic cylinder 31 to establish the initial contact pressure. Subsequently, the excitation generation module 101 executes the active excitation signal generation step. The core logic of this step is to utilize the high-frequency response characteristics of the electro-hydraulic servo valve 32 to convert the electrical signal into hydraulic oil pressure pulsation without changing the macroscopic position of the shearing arm, and then into a mechanical wave acting on the workpiece 40.

[0045] Excitation generation module 101 generates valve port control signal The specific process includes the following sub-steps:

[0046] Determine the basic bias signal. The excitation generation module 101 first acquires the pressure required to maintain the current contact pressure. The required static current or voltage value is defined as the bias signal. The bias signal The specific value depends on the zero-position characteristics of the electro-hydraulic servo valve 32 and the dead zone range of the hydraulic system. The edge controller 100 determines the value by querying a pre-calibrated valve control curve. The voltage must be greater than the dead zone starting voltage of the valve core to ensure that the hydraulic circuit is in a critical conduction state, thereby eliminating the nonlinear hysteresis of signal transmission.

[0047] Synthesize time-varying excitation signals. The excitation generation module 101 uses a bias signal... Based on this, a periodic fluctuation component is superimposed. Valve control signal Follow the mathematical model as follows:

[0048] ;

[0049] in, for The valve control signal (voltage or current) at any given time; Basic bias signal (static operating point voltage / current); The preset excitation amplitude controls the intensity of chatter. The preset detection frequency controls the speed of flutter. The time variable is used. The physical meaning of this formula is to control the valve core to reciprocate slightly near the equilibrium position, so that the instantaneous flow rate entering the shearing hydraulic cylinder 31 fluctuates sinusoidally. Due to the bulk modulus characteristics of hydraulic oil in a closed cavity, this flow rate fluctuation will be directly converted into pressure fluctuation.

[0050] Set excitation parameter constraints. To ensure effective transmission of the probe signal without damaging the device, the edge controller 100 determines the following strategy: and The possible values ​​of:

[0051] For the preset detection frequency The preferred frequency range is 10Hz to 40Hz. This frequency range is chosen based on two factors: firstly, the frequency must be lower than the upper limit of the frequency response bandwidth of the electro-hydraulic servo valve 32 (typically 50Hz-100Hz) to avoid amplitude attenuation of the signal during electro-hydraulic conversion; secondly, the frequency must avoid the inherent frequency set of the device's mechanical structure. The excitation generation module 101 reads the data obtained from the modal analysis experiment from the memory. (For example, the first-order natural frequency is usually distributed between 2Hz and 5Hz), and a safe bandwidth is set. (e.g., 2Hz). Conditions to be met: , This is to prevent structural resonance of the shear arm assembly 30.

[0052] in, Preset detection frequency; It is the set of natural frequencies of a mechanical structure; It represents a certain natural frequency in the set; For safe bandwidth.

[0053] For the preset excitation amplitude Its value is limited to 1% to 5% of the rated control signal amplitude of the electro-hydraulic servo valve 32. This small amplitude setting is based on the fact that it is combined with the enormous inertial load mass of the shear arm assembly 30. Equivalent spring stiffness of hydraulic oil, slight flow pulsation The driving force generated is insufficient to overcome the inertial load and produce macroscopic displacement (i.e., displacement amount). However, this is sufficient to establish a dynamic pressure wave inside the hydraulic cylinder that can be detected by sensors. This constraint ensures that the detection process is a quasi-static non-destructive testing.

[0054] Signal conversion and driving. The excitation generation module 101 converts the calculated digital quantity... The signal is converted into an analog current signal by the digital-to-analog converter unit of the edge controller 100. This analog signal directly acts on the torque motor coil of the electro-hydraulic servo valve 32. In actual operation, if the electro-hydraulic servo valve 32 has a built-in chatter function, the excitation generation module 101 can modify its built-in chatter parameter register by sending a command to adjust the chatter frequency to... Adjust the flutter amplitude to The above-mentioned stimulus is achieved by directly utilizing the underlying hardware functions, thereby reducing the real-time computing load of the edge controller 100.

[0055] After the excitation generation module 101 drives the shearing hydraulic cylinder 31 to generate hydraulic pulsation, the edge controller 100 performs the physical transmission and data acquisition steps of the cross-end signal. This step aims to utilize the attenuation and transmission characteristics of mechanical vibration waves in different media to map the invisible material properties (stiffness or damping) inside the workpiece 40 into measurable kinematic physical quantities at the end of the pressure arm assembly 20.

[0056] The specific process of cross-end signal transmission and acquisition includes the following sub-steps:

[0057] Excitation force conversion and injection. Dynamic pressure wave established inside the shearing hydraulic cylinder 31. Effective working area of ​​the piston acting on the shearing hydraulic cylinder 31 The upward force is transformed into alternating mechanical driving force. The driving force is directly coupled to the contact surface of the workpiece 40 through the blade tip of the shear arm assembly 30. In this process, since the mechanical structural stiffness of the shear arm assembly 30 is much greater than the structural stiffness of the workpiece 40 to be disassembled (such as a scrapped car body), the elastic deformation of the shear arm itself is regarded as a high-order infinitesimal quantity and is negligible. It is approximately assumed that the hydraulic energy generated by the electro-hydraulic servo valve 32 is completely converted into mechanical excitation energy acting on the input end of the workpiece 40.

[0058] Medium transmission and impedance modulation. The mechanical wave propagates inside the workpiece 40 and eventually reaches the clamping interface of the pressure arm assembly 20. In this process, the workpiece 40 acts as a complex mechanical impedance element, and its modulation of the vibration signal follows a single-degree-of-freedom dynamic equilibrium equation:

[0059] ;

[0060] in, The output force (receiving end) is transmitted to the contact interface of the pressure arm assembly. The input force applied to the shear arm; The equivalent mass of the workpiece; The equivalent damping coefficient of the workpiece; The equivalent structural stiffness of the workpiece; , , These represent the displacement, velocity, and acceleration generated by the workpiece under excitation, respectively. This formula indicates that when the workpiece 40 is a high-rigidity component (such as...), When the value of the A-column or main beam is large, its structural deformation The smaller the energy, the less energy is absorbed or attenuated by the workpiece itself, and the less force is transmitted to the output end. The amplitude is relatively large; conversely, if the workpiece 40 has a loose structure, the signal attenuation is significant.

[0061] Receiver dynamic response and signal pickup. Output force transmitted to pressure arm assembly 20. The pressure arm assembly 20 is driven to generate a small-amplitude forced vibration. At this time, the pressure arm assembly 20 is physically equivalent to a second-order oscillating system consisting of mass, spring, and damping. Here, the mass is the equivalent moving mass of the pressure arm assembly 20. The spring stiffness is mainly determined by the hydraulic stiffness of the closed oil column in the rodless chamber of the hydraulic cylinder 21. supply.

[0062] hydraulic stiffness Determined by the following formula:

[0063] ;

[0064] in, The hydraulic stiffness of the hydraulic cylinder for the pressure arm; This refers to the effective working area of ​​the piston in the hydraulic cylinder for the pressure arm. This refers to the total volume of the pressure chamber currently being pressurized in the hydraulic cylinder of the pressure arm; This is the effective bulk modulus of the hydraulic oil. This parameter is affected by the proportion of air mixed in the oil and the expansion of the hydraulic hoses, and is usually taken as a value of [value missing]. to .

[0065] Vibration acceleration collected by accelerometer 23 It depends not only on the input force It is also subject to the reaction force of the pressure arm's own inertia, and its dynamic relationship is as follows:

[0066] ;

[0067] in, The equivalent moving mass of the pressure arm assembly; The acceleration at the end of the pressure arm is collected by an accelerometer. This is the damping coefficient of the pressure arm motion; The speed of the pressure arm movement; For hydraulic stiffness; For the micro-displacement of the pressure arm; This is the driving force transmitted to the pressure arm.

[0068] By keeping the position of the hydraulic cylinder 21 of the pressure arm locked, the system , and All of these can be considered as known constants. Therefore, the measured acceleration Force transmitted from the workpiece They exhibit a positively correlated linear mapping relationship.

[0069] Signal digitization and sampling constraints. The edge controller 100 performs analog-to-digital conversion on the received analog voltage signal. This is done to fully preserve the excitation frequency. Based on its characteristic information, the data acquisition channel of the edge controller 100 is configured to follow the Nyquist sampling theorem. Specifically, the sampling frequency... Set as detection frequency More than 10 times (i.e.) The sampling frequency should be no less than 200Hz. For example, when the excitation frequency is 20Hz, the sampling frequency is set to 1kHz. The effect of this high-rate sampling technique is to eliminate aliasing errors while ensuring the analytical accuracy of phase information, thereby accurately calculating the phase lag between the input and output signals and assisting in judging the damping characteristics.

[0070] The feature extraction module 102 performs the processing step of converting the time-domain discrete sampled signal into frequency-domain spectral data. The physical essence of this step is to utilize the frequency selection characteristics of frequency domain analysis to orthogonally decompose the mixed signal containing ambient broadband noise (such as engine vibration and hydraulic pump pulsation), thereby accurately separating the signal with frequencies of [frequency range missing] under complex field conditions. The active detection component.

[0071] Feature extraction module 102 targets the input pressure signal of shear arm assembly 30. and the output acceleration signal of the pressure arm assembly 20 The following sub-steps will be executed simultaneously:

[0072] Data buffering and windowing modulation. The edge controller 100 maintains a dual-channel circular buffer in memory to store sampling data points uploaded by the pressure sensor 33 and the accelerometer 23 in real time. When the amount of new data in the buffer reaches the preset calculation window length... At that time, the frequency domain solution task is triggered. To balance spectral resolution and the real-time response speed of the control system, the window length is... The selection of [a] must comply with the following constraints:

[0073] ;

[0074] in, This refers to the system sampling frequency; To calculate the window length (number of data points); The minimum frequency resolution required by the system is typically set to 0.5Hz to 1Hz. Meanwhile, to ensure that the update cycle of control commands is less than the system's mechanical constant, The preferred value range is 512 to 2048 points.

[0075] Feature extraction module 102 extracts the truncated data. Point time-domain sequence A windowing function is applied to suppress spectral leakage caused by non-integer period sampling. The Hanning window is used as the modulation function, and the windowed sequence... The calculation is as follows:

[0076] ;

[0077] in, The time-domain sequence after windowing; This is the original time-domain sampling sequence; For sampling point index, Although the introduction of the window function will result in some signal energy loss (widening of the main lobe), it can significantly reduce the sidelobe amplitude and prevent strong background noise energy from leaking to the detection frequency. Within the frequency band in which it is located, this improves the signal-to-noise ratio.

[0078] Fast Discrete Fourier Transform. Feature extraction module 102 uses a radix-2 decimation-time algorithm to extract the windowed time-domain sequence. Convert to frequency domain complex sequence The transformation process follows the mathematical model below:

[0079] ;

[0080] in, It is a sequence of complex numbers in the frequency domain; For frequency spectral line index, ; The unit is the imaginary number. This operation is performed in the DSP coprocessor or floating-point unit of the edge controller 100, utilizing hardware acceleration to control the time of a single transformation to the microsecond level.

[0081] Target spectral line locking and amplitude normalization. To obtain the true signal intensity with physical dimensions, the feature extraction module 102 does not calculate the full-band spectrum, but directly locks the signal corresponding to the detection frequency. Specific spectral index :

[0082] ;

[0083] in, This is the index of the discrete spectral lines corresponding to the detection frequency.

[0084] Subsequently, the physical amplitude at that frequency point is calculated based on the energy recovery coefficient of the window function. :

[0085] ;

[0086] in, To detect the physical amplitude at the frequency; The selected sequence of window functions; For frequency domain sequences in The modulus at that point; Let be the coherence gain coefficient of the window function. For the Hanning window, Coefficient 22 is used to convert the energy of the bilateral spectrum to the unilateral spectrum.

[0087] Through the above processing, the feature extraction module 102 outputs two key scalars: the input pressure amplitude. (Unit: MPa) and output acceleration amplitude These two values ​​represent the energy components that actually exist in the system and are actively excited by the electro-hydraulic servo valve 32 after eliminating all background interference.

[0088] Impedance identification module 103 performs a signal-to-noise ratio (SNR) confidence check. This step is based on the principle of the difference in transmission efficiency of mechanical waves at different medium interfaces: when the shear arm assembly 30, workpiece 40, and pressure arm assembly 20 form a rigid metal closed loop, the transmission loss of vibration energy at a specific frequency is minimal, and the received signal is significantly higher than the background noise; however, when there are high-damping non-metallic materials such as rubber or plastic in the link, or when there are mechanical gaps, the high-frequency vibration energy is dissipated or interrupted, and the received signal is submerged in the background noise. The system utilizes this physical characteristic to quantify the effectiveness of physical coupling by calculating the frequency domain SNR.

[0089] The specific sub-steps executed by the impedance identification module 103 are as follows:

[0090] Energy integration within the signal band. Impedance identification module 103 calculates the main lobe energy of the detected signal based on the frequency domain power spectrum data output by feature extraction module 102. To encompass the frequency spread energy caused by system damping, the integration interval is set to the detection frequency. The neighborhood centered on the signal. Total signal power. The calculation formula is as follows:

[0091] ;

[0092] in, The total power of the main lobe of the detection signal; The magnitude of the frequency domain sequence; This refers to the discrete spectral line index corresponding to the detection frequency; This is the half-bandwidth index span. The value is determined by the calculation window length. and sampling frequency Decision, that is In a preferred embodiment, the signal integration bandwidth... Set to 2Hz to ensure coverage of the main frequency and its sidelobe leakage energy.

[0093] Background noise floor estimation. Impedance identification module 103 at the detection frequency. Noise reference frequency bands are selected on both sides. To avoid interference from spectral leakage of the probe signal in noise estimation, and to avoid hydraulic pulsation interference in the low-frequency band (<10Hz), the selection of the noise reference frequency band follows the following frequency constraints:

[0094] ;

[0095] in, This is a set of noise reference frequency bands; The protection interval is set to 3Hz to 5Hz to isolate the main lobe of the signal; The noise sampling range is set to 10Hz to 20Hz.

[0096] Impedance identification module 103 calculates the average noise power density within the reference frequency band. :

[0097] ;

[0098] in, This represents the average noise power density. For set The total number of valid spectral lines included. The physical significance of this step is to extract the environmental vibration benchmark under the current operating conditions in real time, eliminating the influence of environmental noise fluctuations caused by changes in engine speed or hydraulic pump load on the judgment.

[0099] Signal-to-noise ratio (SNR) calculation and threshold determination. Impedance identification module 103 calculates the dimensionless SNR index. :

[0100] ;

[0101] in, This is a frequency domain signal-to-noise ratio indicator.

[0102] Then, Compared with the preset confidence threshold Compare them. The determination is made using the following calibration method: During the equipment factory commissioning phase, "no-load operation" is tested respectively. (No signal at the location) and "clamping standard rigid steel ingot" ( Two states (full signal) are recorded. The maximum residual ratio under no-load conditions is recorded. The minimum ratio to the rigid clamping state ,Pick Under typical operating conditions, The value range is from 5.0 to 8.0.

[0103] Coupled state logic output. Impedance identification module 103 outputs a verification flag based on the comparison result:

[0104] when When the physical coupling confidence level is high, workpiece 40 has rigid transmission characteristics, allowing it to proceed to the subsequent stiffness calculation process.

[0105] when When this condition is met, the physical coupling confidence level is determined to be low. This state indicates that the current vibration response is not significantly higher than the ambient noise, physically corresponding to situations such as the workpiece being made of energy-absorbing material, workpiece breakage, or hydraulic cylinder not being compacted. Based on this, the system triggers safety logic, prohibiting the output of high-pressure back pressure commands to prevent malfunctions under non-rigid operating conditions.

[0106] After passing the signal-to-noise ratio confidence test, the impedance identification module 103 performs the equivalent stiffness index test. The quantitative calculation steps are based on the principle of vibration transmission efficiency: in a series mechanical system, the greater the stiffness of the intermediate medium, the smaller the transmission loss of high-frequency dynamic forces, and the higher the response energy at the receiving end; conversely, if the intermediate medium is an energy-absorbing material or has a loose structure, the dynamic force will be significantly attenuated. The system constructs a quantitative index characterizing the material properties of the workpiece by calculating the ratio between the effective driving force at the input end and the inertial response force at the output end.

[0107] The solution process specifically includes the following sub-steps:

[0108] Effective excitation force calculation. Impedance identification module 103 calculates the input pressure amplitude output by feature extraction module 102. Calculate the effective dynamic excitation force acting on the piston rod end of the shear arm assembly 30. Because of dynamic sealing friction and viscous damping during the micro-motion of the hydraulic cylinder, this portion of energy is not transferred to the workpiece and must therefore be subtracted from the frequency domain amplitude calculation. Effective dynamic excitation force Calculated by the following formula:

[0109] ;

[0110] in, This represents the effective dynamic excitation force amplitude in the frequency domain. Input pressure amplitude; The effective pressure-bearing area of ​​the shearing hydraulic cylinder 31 is taken as the area of ​​the rodless cavity when the control logic is for the push-out excitation and the area of ​​the rod cavity when the retraction excitation is taken. The preset constant is determined by applying the same frequency when the equipment is in an unloaded state. The flutter excitation was measured, and the minimum pressure amplitude required to maintain the movement of the hydraulic cylinder was measured and converted into a force value.

[0111] Mechanical transmissibility calculation. The impedance identification module 103 calculates the transmissibility of the mechanical wave after it passes through the workpiece 40 based on the dynamic response of the pressure arm assembly 20. In this physical model, the pressure arm assembly 20 is considered as a signal receiver, and the acceleration it generates... It is the result of the transmitted force driving its inertial mass. Transmissibility The definition is as follows:

[0112] ;

[0113] in, Mechanical transmission rate; To output the acceleration amplitude; The equivalent inertial mass of the pressure arm assembly 20 is equal to the sum of the "pressure arm mechanical structure mass" and the "equivalent hydraulic oil mass in the pressure arm hydraulic cylinder". This parameter is calibrated by the equipment manufacturer through modal testing and then fixed in the controller. The range is theoretically The closer the value is to 1, the closer the workpiece's transmission of vibration is to that of a rigid body.

[0114] Stiffness exponent normalization mapping. To adapt to the lookup logic of the collaborative calculation module 104, the impedance identification module 103 uses the physical transmissivity... Mapped to dimensionless standardized stiffness exponent (Range 0-100). The calculation formula is as follows:

[0115] ;

[0116] in, The standardized equivalent stiffness index; and The calibration parameters for the system must adhere to strict operating condition constraints to ensure consistency.

[0117] 1. (High rigidity benchmark): When the hydraulic system is at its rated operating temperature (e.g., 40℃-60℃) and preset contact pressure... The transfer rate measured under the condition of clamping a standard solid steel ingot (e.g., 45# steel, with the same dimensions as a typical workpiece);

[0118] 2. (Low Stiffness Reference): Transmission rate measured under the same temperature and pressure conditions, while clamping a standard high-damping rubber block (Shore hardness 60A-70A). Through dual-point calibration, the system can automatically compensate for drift in transmission characteristics caused by changes in hydraulic oil viscosity or mechanical wear.

[0119] Timing smoothing filtering. Considering the instantaneous impact interference at the disassembly site (such as drops from other parts of the vehicle body), the single-step calculation... There may be noise. The impedance identification module 103 introduces a first-order low-pass filter algorithm to smooth the stiffness exponent of the continuously output:

[0120] ;

[0121] in, This is the final stiffness exponent after filtering; For standardized equivalent stiffness exponent (instantaneous value); For discrete-time ticks; This is the filtering smoothing coefficient. The value range is set to 0.2 to 0.4. The engineering basis for selecting this parameter is to filter out random fluctuations with a frequency higher than 5Hz, and to ensure that the response delay of the control system to changes in workpiece stiffness does not exceed 200ms, thereby meeting the requirements of real-time adaptive control.

[0122] The collaborative computing module 104 calculates the equivalent stiffness index based on the previous steps. The adaptive flow rate speed regulation step is then executed for the shear arm assembly 30. The physical essence of this step is load adaptive adjustment based on the principle of "constant power control." This is because the output power of the hydraulic power source (engine and hydraulic pump) has a physical upper limit. When the workpiece is identified as having extremely high stiffness (corresponding to high shear resistance and high system pressure), When the flow rate is high, the control system must actively reduce the flow rate. This is to prevent the total power demand from exceeding the engine load capacity and causing engine stalling, while also reducing the throttling heat loss caused by high-pressure fluid flowing through the relief valve.

[0123] The specific implementation process of shear arm flow feedforward control includes the following sub-steps:

[0124] Operation command parsing and normalization. The collaborative computing module 104 receives the raw control signals sent by the operating handle. To eliminate mechanical jitter near the center position of the handle potentiometer, the module sets a dead-time threshold (e.g., ±5%). When the signal amplitude exceeds the dead-time, it is linearly mapped to a normalized flow demand command. Its value range is [0,1].

[0125] Stiffness-flow-gain mapping coefficient calculation. Cooperative calculation module 104 uses the equivalent stiffness index. Calculate the flow correction gain coefficient In order to achieve a smooth and rapid switching between the two working conditions of "efficient shearing of soft materials" and "strong shearing of hard materials", this embodiment abandons the simple linear lookup table and adopts a nonlinear decay model based on the S-shaped function. The calculation formula is as follows:

[0126] ;

[0127] in, Adjust the gain coefficient for flow rate; , Minimum / maximum flow gain; The sensitivity factor for the sigmoid function; This represents the stiffness value at the inflection point of the working condition.

[0128] The definitions and determination methods for each parameter are as follows:

[0129] (Maximum flow gain): For low stiffness conditions, the value is 1.0, indicating that full-speed operation is allowed.

[0130] (Minimum Flow Gain): Corresponds to the ultimate stiffness condition. This value must meet certain conditions. ,in For system overflow pressure, The rated power of the engine. This is the effective power utilization factor. A typical value ranges from 0.3 to 0.4, meaning the flow rate is limited to 30%-40%.

[0131] (Inflection Point): Defines the critical stiffness value at which the flow rate begins to significantly decrease, typically set to... .when At that time, the system determines that the workpiece has entered the "difficult-to-machine material" range.

[0132] (Sensitivity factor): Determines the steepness of the operating condition transition. The larger the value, the more sensitive the system is to changes in stiffness, and the more drastic the flow rate switching. The preferred value range is 0.1 to 0.2. This range is selected to ensure that the flow rate gain can smoothly decrease from 0.9 to 0.4 within the transition range of stiffness index from 40 to 60, avoiding hydraulic shocks caused by sudden changes in flow velocity.

[0133] Pressure saturation correction. Based on stiffness-based feedforward control, the collaborative calculation module 104 incorporates the current shear pressure. As a safety constraint variable, the pressure saturation coefficient is calculated. This logic serves as a fallback protection mechanism in the event of "stiffness identification failure". The calculation is as follows:

[0134] ;

[0135] in, This is the pressure saturation coefficient; This represents the current actual shear pressure; Set the opening pressure of the main relief valve of the hydraulic system (e.g., 35MPa). The threshold for mandatory traffic intervention is set to [value]. 85% to 90% (e.g., 30 MPa). When the actual shear pressure approaches the safety limit, regardless of the stiffness identification result, the formula forces the flow command to zero, prioritizing the maintenance of system pressure establishment to prevent the relief valve from being open for a long time, which would cause the oil temperature to overheat.

[0136] The final valve control command is generated. The collaborative calculation module 104 integrates the above calculation results and generates a valve opening command to be sent to the electro-hydraulic servo valve 32. :

[0137] ;

[0138] in, This is the valve opening and flow rate command; The nominal flow rate of the electro-hydraulic servo valve 32 at the rated pressure drop (e.g., a single-sided pressure drop of 3.5 MPa); This is a command required for normalizing the controller.

[0139] Dead-time compensation and drive output. Before converting the command into drive current, the edge controller 100 adds a dead-time compensation component. Final output current. The calculation is as follows:

[0140] ;

[0141] in, For output current; A lookup function for the valve's flow and current characteristics; To compensate for dead zone current. The determination method is as follows: In offline calibration mode, gradually increase the coil current at a slope of 0.1mA / s until the laser displacement sensor detects a 1% displacement of the valve core; this current value is the value of the current. This compensation ensures that... Even at extremely low speeds (such as micro-feeding during high-hardness shearing), the control signal can still overcome the static friction of the valve core, achieving precise pressure control without hysteresis.

[0142] While executing the flow control of the shear arm, the collaborative computing module 104 simultaneously performs the active energy absorption control steps for the pressure arm assembly 20. This step is based on the principle of variable impedance control: using electro-hydraulic proportional technology to adjust the back pressure of the rodless chamber of the pressure arm hydraulic cylinder 21 in real time, making it behave as a dynamic spring system with variable stiffness and damping. When the workpiece breaks and releases huge elastic potential energy, the system instantaneously increases the hydraulic resistance, forcing the hydraulic oil to do work during the overflow process, thereby converting the rebound kinetic energy of the workpiece into heat energy dissipation and preventing destructive impact on the mechanical structure.

[0143] The specific implementation process of active energy absorption control of the pressure arm includes the following sub-steps:

[0144] The reference clamping pressure is set. The collaborative calculation module 104 first calculates based on the equivalent stiffness index. Calculate the static reference pressure required for hydraulic cylinder 21 of the pressure arm. This setting follows the principle of "stiffness matching," meaning that the applied clamping force should be compatible with the workpiece's compressive strength to ensure locking friction while avoiding crushing the workpiece. The calculation formula is as follows:

[0145] ;

[0146] in, This is the static reference pressure of the pressure arm; This is the maximum safe pressure for the hydraulic system. The rated working pressure of the hydraulic cylinder for the pressure arm; To maintain the minimum preload pressure for clamping contact (note the distinction from the minimum pressure in subsequent fault logic); This represents the material hardening asymptotic factor.

[0147] Dynamic damping pressure synthesis. To suppress high-frequency flutter during shearing, the collaborative computing module 104 constructs an electronic damping loop based on velocity feedback. First, the signal acquired by the accelerometer 23 is processed. Integrate to obtain the vibration velocity To eliminate the effects of integral accumulation error and gravitational components, a cutoff frequency is connected in series in the system. Hz digital high-pass filter:

[0148] ;

[0149] in, To filter out vibration velocity after low-frequency drift; For high-pass filter operators; This is the raw acceleration signal acquired in real time by the accelerometer (sensor 23). This signal typically combines high-frequency mechanical vibration, low-frequency mechanical motion, gravitational components, and the sensor's zero-point drift error.

[0150] Subsequently, the dynamic damping pressure compensation value was calculated. The physical meaning of this compensation value is that when the pressure arm experiences severe swaying (high speed), it automatically increases the opening pressure of the relief valve, increasing the resistance to hydraulic oil discharge, thereby producing an energy absorption effect similar to a viscous damper. The calculation formula is as follows:

[0151] ;

[0152] in, This is the dynamic damping pressure compensation value; This is the virtual damping coefficient (unit: MPa / (m / s)). The determination of the critical damping state of the system needs to be referenced, and its calculation formula is: .in For the preset damping ratio, a value of 0.7 to 0.9 (overdamped) is used for rigid workpieces, and a value of 0.4 to 0.6 (underdamped) is used for flexible workpieces. For hydraulic stiffness; For the mass of the pressure arm; This refers to the piston area of ​​the hydraulic cylinder for the pressure arm.

[0153] Fracture prediction and impact interlocking. The collaborative computing module 104 monitors the input pressure of the shear arm in real time. Time differential When detected Below the preset pressure drop threshold At that time, it is determined that the workpiece is about to undergo brittle fracture. The calibration method is as follows: During the system debugging phase, control the shear arm to run at full speed under no-load and then suddenly stop it, and record the maximum pressure drop slope at this time. ,Pick (e.g., -40MPa / s to -60MPa / s) to distinguish between normal unloading and fractured unloading.

[0154] Once the fracture detection is triggered, the system's cumulative duration is... (150ms) pulse lock-in pressure The final total pressure command generated. for:

[0155] ;

[0156] in, This is the total back pressure command sent to the proportional relief valve; The pulse lockout pressure; The time derivative of the shear pressure; The pressure drop threshold; This is an indicator function; it is 1 if the condition is met, and 0 otherwise. The pressure is typically set between 5 MPa and 10 MPa. The purpose of this pulse pressure is to instantly "harden" the hydraulic cylinder of the pressure arm, utilizing the incompressible properties of hydraulic oil to resist the enormous rebound force at the moment of workpiece fracture.

[0157] Command translation and chatter drive. The collaborative computing module 104 translates the total pressure command... Convert the control current of the proportional solenoid valve by referring to the table. To reduce the static friction and hysteresis of the valve core, the edge controller 100 superimposes a high-frequency chatter signal onto the DC control current. The chatter signal is defined as a triangular wave with a frequency of 150Hz and an amplitude of 5% of the rated current. This process significantly improves the relief valve's response sensitivity to minute pressure adjustment commands, ensuring the real-time performance of damping control.

[0158] While executing the aforementioned adaptive control strategy, the collaborative computing module 104 continuously runs system state monitoring and fault-safe logic in the background. This logic is based on the principle of "deterministic state degradation": when the reliability of sensor feedback decreases or the control algorithm becomes numerically unstable, the system must immediately disconnect the adaptive closed loop that may generate positive feedback, forcing the control strategy to converge to a pre-verified, energy-constrained open-loop or semi-closed-loop parameter set. This mechanism ensures that under any single hardware failure or software anomaly, the output energy of the hydraulic system is always confined within non-destructive physical boundaries.

[0159] The specific implementation process of fault-oriented safe mode includes the following sub-steps:

[0160] Multidimensional signal integrity verification. The edge controller 100 periodically scans the input pressure signal with the highest interrupt priority. Output acceleration signal and control handle commands The verification logic constructs a fault state vector by comparing the electrical and physical characteristics of signals. The system defines fault triggering conditions. The system is deemed to have entered a fault state when any of the following logical conditions are met:

[0161] ;

[0162] in, This is a fault status indicator; The voltage sampled by the sensor; and For electrical boundaries; The rate of change of the signal; It is the physical gradient limit; For flow instructions; This is the instruction threshold; To provide feedback on pressure; This is the minimum pressure build-up threshold for the hydraulic system.

[0163] The physical meaning and parameters of each condition are determined as follows:

[0164] Electrical range exceeded: The sensor sampling voltage, (0.2V) and (4.8V) is the threshold for determining whether the circuit is open or short-circuited.

[0165] Physical gradient anomaly: The rate of change of the signal; This represents the maximum allowable response speed of the physical system. For hydraulic pressure signals, Set to 1.5 times the pump source response bandwidth (e.g., 200 MPa / s); for acceleration signals, Set to 10% of the sensor's range per sampling period. Any change exceeding this rate is considered electromagnetic interference or poor contact.

[0166] Logical interlock conflict: The third term represents the "pipe burst" or "complete pressure loss" condition, i.e., when the flow command... Greater than the threshold (50%) and the duration exceeds Feedback pressure at (300ms) Still below the minimum pressure threshold (2MPa).

[0167] The shear arm control strategy degrades. When... When the confidence flag of the position or stiffness calculation module is invalid, the collaborative calculation module 104 immediately disables the stiffness-based flow feedforward control logic. The system switches to "manual speed limiting mode". In this mode, the flow command of the electro-hydraulic servo valve 32... Decoupled from sensor feedback, responding only to manual commands, and with added safety clamping:

[0168] ;

[0169] in, This is a safe mode traffic command; The safety degradation factor ranges from 0.3 to 0.5. This parameter is set to limit the maximum speed of the hydraulic cylinder, ensuring that the kinetic energy generated is insufficient to damage the cab safety net or cause the vehicle to overturn in case of misoperation, thus providing operators with sufficient reaction and correction time.

[0170] The arm control strategy is downgraded. If the acceleration sensor 23 or the pressure sensor 33 fails, the collaborative computing module 104 immediately disables the active energy absorption control logic. The system switches to "constant pressure holding mode". In this mode, the closed-loop target value of the arm hydraulic cylinder 21 is locked at a fixed safe pressure. :

[0171] ;

[0172] The default pressure The selection principle follows the "minimum effective clamping" standard: the clamping force it generates is greater than the frictional requirements caused by the weight of the workpiece, but less than the yield strength of conventional metal structural parts, ensuring that the workpiece will not fall off in blind operation, nor will it crush objects of unknown stiffness due to excessive pressure.

[0173] Closed-loop stability monitoring and oscillation suppression. During adaptive control operation, if control parameters become mismatched due to sudden changes in workpiece stiffness or nonlinear factors, the system may generate self-excited oscillations. The collaborative calculation module 104 calculates the energy product of the control deviation through a sliding window. To monitor stability:

[0174] ;

[0175] in, To control the deviation energy product; The observation window length (e.g., 200ms); To control deviation; This is the full-scale value of the corresponding physical quantity (used for normalization). If Exceeding the stability threshold (For example, 0.05, meaning the average deviation energy exceeds 5% of full scale), the system determines that the control is diverging. At this time, the system performs a "parameter annealing" operation, removing all dynamic gain coefficients (such as...) , Multiply by the decay factor (Take 0.5):

[0176] ;

[0177] in, For the next discrete control cycle ( The updated dynamic gain coefficient at time (time) is a new parameter value after annealing and numerical reduction, used to suppress system oscillations in the next time step; For the current discrete control period ( The dynamic gain coefficient currently in use (at any given time). This symbol is a wildcard and represents a specific control parameter in actual operation, such as flow correction gain. or virtual damping coefficient ; The parameter annealing attenuation factor (a dimensionless scalar) is set to 0.5 in this paper. This coefficient is used to force the control gain to be halved when control divergence is detected, thereby sacrificing the system's response speed in exchange for stability margin. This is a discrete-time series index, representing the current control calculation cycle.

[0178] This operation is physically equivalent to instantaneously increasing the system's damping ratio, forcibly dissipating the oscillating energy within the system by sacrificing the agility of the dynamic response, until the system returns to a steady state.

[0179] Human-Machine Interface Warning. While implementing the degradation strategy, the edge controller 100 sends specific fault codes to the human-machine interface terminal via the fieldbus. The fault codes include the fault source (sensor ID), fault type (electrical / physical / logical), and current operating mode (speed limit / pressure control). For critical faults involving structural safety (such as pipe burst detection), the controller will directly cut off the power supply circuit to the pilot solenoid valve, physically locking all hydraulic actuators until the system restarts or the fault is cleared.

[0180] After completing impedance identification and control parameter calculation, the collaborative computing module 104 within the edge controller 100 performs parameter preloading and timing synchronization. This step aims to resolve the time-domain mismatch between the "high-speed electronic control system" and the "low-speed, high-inertia hydraulic system." Because it takes time for the solenoid valve coil to establish its magnetic field (electromagnetic lag) and for the hydraulic oil to push the piston through the pipeline to build pressure (transmission and pressure build-up lag), there is an objective phase difference between the issuance of control commands and the execution of physical actions. Without compensation, the adaptive parameters (such as stiffness matching flow rate) calculated based on the current sensor data will always lag behind the actual deformation state of the workpiece. This embodiment achieves phase lead synchronization of control commands through a "dynamic hysteresis estimation" and "shadow register preloading" mechanism.

[0181] The specific implementation process of timing synchronization and parameter preloading includes the following sub-steps:

[0182] Multi-source sensor timestamp alignment. The edge controller 100 utilizes a hardware timer within the FPGA to provide a unified global time base for all acquisition channels. Since different sensors have different response bandwidths (e.g., 1kHz for a pressure sensor and 5kHz for an accelerometer), the collaborative computing module 104 establishes a time alignment buffer. The module uses the highest sampling frequency as a reference and performs zero-order hold or linear interpolation on the low-frequency signal to generate a time stamp with a unified time. state vector This ensures that all physical quantities involved in the calculation are strictly aligned on the time axis.

[0183] Real-time estimation of system physical lag time. The collaborative calculation module 104 calculates the total response lag time of the hydraulic system in real time based on the current operating conditions. This hysteresis time is not a fixed value, but rather a nonlinear function of hydraulic oil temperature and system pressure. Low temperature increases oil viscosity, which prolongs the hysteresis due to increased frictional resistance in the pipeline; low pressure causes microbubbles mixed in the oil to expand, reducing the bulk modulus and thus prolonging the pressure build-up time. The calculation model is as follows:

[0184] ;

[0185] in, Total response lag time; This is due to electrical lag; The mechanical reference lags behind; This refers to the viscosity-temperature characteristic coefficient; , , Reference temperature and actual oil temperature; The gas content influence coefficient; Current system pressure; To prevent small quantities from being excluded.

[0186] The definitions and determination methods for each parameter are as follows:

[0187] (Electrical hysteresis): The time it takes for the solenoid valve coil current to rise from 0 to 63.2% of the operating current, which is typically 10ms to 15ms (determined by the inductance).

[0188] (Mechanical reference hysteresis): at the reference temperature The time from valve core displacement to hydraulic cylinder action, measured at (50℃) and rated pressure, is typically 30ms to 40ms.

[0189] For real-time measurement of hydraulic oil temperature, The viscosity-temperature characteristic coefficient (taken as 0.015 to 0.025);

[0190] The current main pressure, The gas content influence coefficient (ranging from 2.0 MPa to 5.0 MPa) is used to compensate for cavitation hysteresis under low pressure. To prevent small amounts (0.1 MPa) with a denominator of zero.

[0191] Phase lead step size calculation and parameter preloading. The collaborative computing module 104 calculates the required phase lead step size. To compensate for physical lag. This indicates how many calculation cycles in advance the control system needs to prepare the parameters to ensure that the parameters take effect precisely at the moment of the physical action:

[0192] ;

[0193] in, This is the phase lead step size (an integer value). It indicates that in order to compensate for physical lag, the control algorithm needs to "predict" the parameters calculated at the current moment as the effective parameters for the next Nahead-Nahead cycle. This is a rounding function. It is used to convert the time ratio of a floating-point number into the number of discrete periods. This is the total physical response lag time of the system estimated in real time. This time includes the total delay of electrical transmission, mechanical action, and hydraulic pressure build-up; The control cycle period (i.e., the sampling and calculation step size of a discrete control system, such as 1ms) is the time base unit of a digital controller. To control the cycle period (e.g., 1ms), the collaborative computing module 104 will calculate the control parameters (such as flow gain). Damping coefficient The data is written to the microprocessor's "shadow register." The shadow register is a buffer storage area connected in parallel with the DAC output register. At the update time of each PWM cycle, the direct memory access controller automatically and atomically moves the data in the shadow register to the active register. This mechanism allows the software algorithm to calculate parameters at any time, while the hardware driver only updates the output on the specified clock edge, thereby eliminating waveform distortion caused by jitter due to computation time.

[0194] Timing handshake and validity window verification. The system performs a validity handshake verification before writing data to the shadow register. The collaborative computing module 104 checks the sensor data timestamps on which the current parameter calculation is based. With current system time and predicted effective time The relationship between them. The logic for determining validity is as follows:

[0195] ;

[0196] in, This is the current system absolute time; The timestamp of the raw sensor data referenced for calculating the current control parameters (i.e., the moment the data was acquired into the system); This is the total physical response lag time of the system estimated in real time; The inherent standard time consumption is preset or predicted in the algorithm model; The allowed time jitter window threshold is set (e.g., 2ms). If the verification passes, it means that the calculated parameters are "fresh" in time and adapted to future physical actions, and writing to the shadow register is allowed; if the verification fails (e.g., due to excessive calculation delay caused by interrupt conflict), the current calculation result is discarded, the safety parameters of the previous cycle remain unchanged, and the timing synchronization warning flag is set.

[0197] The interaction between the collaborative computing module 104 and the underlying driver unit adopts an asynchronous producer-consumer handshake mode based on a state machine. This mode aims to resolve the contradiction between the uncertain time consumption of algorithm computation and the strict periodicity requirements of PWM driving. The algorithm layer, as the producer, generates control parameters at irregular intervals; the driver layer, as the consumer, must read the parameters within a fixed time window. By constructing critical sections for atomic operations, the system ensures strict data consistency of multi-dimensional control variables (flow rate, pressure, damping) at the moment of parameter transmission, preventing data tearing phenomena such as "flow rate command updated but pressure command still old value".

[0198] The specific implementation process of the handshake trigger logic includes the following sub-steps:

[0199] Algorithm layer validity self-check and submission. After completing the strategy calculation for the current cycle, the collaborative computing module 104 first performs a physical feasible region check. The check logic examines the calculated control variables. Does it meet the following constraints:

[0200] ;

[0201] in, The output control quantity is calculated by the collaborative computing module in the current control cycle; This is the physical lower limit or safety lower limit of the actuator (e.g., zero flow, minimum dead zone current, or minimum holding pressure). This refers to the physical or safety limits of the actuator (e.g., maximum physical travel, system overflow pressure, or maximum output current of the driver chip). The change in the control quantity calculated for the current period relative to the previous period; For logical AND operator; This is the maximum permissible rate of change based on the frequency response characteristics of the hydraulic system (e.g., amplitude change corresponding to a 50Hz bandwidth). If the verification passes, the module writes the data to the shadow buffer and sets the data ready flag. Set. The shadow buffer is a memory region physically isolated from the actual driver output register, allowing the algorithm layer to perform write operations without interfering with the current output.

[0202] Phase synchronization monitoring at the driver layer. The underlying driver unit monitors the counter state of the PWM generator in real time. To prevent waveform distortion caused by changing the duty cycle during the high level of the PWM waveform, the handshake operation is strictly limited to the "zero crossing moment" of the PWM cycle. The driver unit generates a synchronization trigger signal. :

[0203] ;

[0204] in, Synchronization trigger signal (logic Boolean value); For indicator functions; For logical OR operators; This is the current counter value of the PWM generator (a real-time variable that increases or decreases linearly over time). Sets a constant value for the period register of the PWM generator. This value determines the frequency of the PWM carrier (e.g., if the clock frequency is fixed, the larger this value, the lower the frequency).

[0205] Only when The duty cycle register can only be updated when the current condition is met. Furthermore, the driver layer must also confirm the current bus current. The short-circuit protection threshold (e.g., 3A) has not been exceeded, and the temperature of the driver chip is within acceptable limits. Below the thermal shutdown threshold (e.g., 125°C).

[0206] Atomic-level double-buffered switching. When algorithm layer data is detected to be ready ( And the driving layer phase synchronization ( When a direct memory access controller (DMA) or interrupt service routine triggers an atomic double-buffered switch operation:

[0207] ;

[0208] This operation is achieved by swapping memory pointers to the two buffers, taking only a few clock cycles (nanoseconds). After the pointer swap, the original shadow buffer becomes the current active buffer, directly mapped to the hardware output; the original active buffer becomes the new shadow buffer, ready for the next computation. This mechanism ensures that regardless of the size of the control parameters, their updates are instantaneous relative to the drive cycle.

[0209] Timeout safety attenuation strategy. If a calculation timeout or communication interruption results in a continuous attenuation... Failed to trigger handshake within one PWM cycle (i.e., (If the value remains at 0), the system enters soft hold mode. To prevent integral saturation or mechanical drift caused by maintaining a fixed value for an extended period, the drive output... Automatic decay is performed according to the following formula:

[0210] ;

[0211] in, For the current discrete control cycle (the 1st) The drive output value of the cycle (usually the PWM duty cycle or target current value sent to the driver chip); For the previous discrete control cycle (the first) The drive output value of the beat; This is the index of the current computational beat in the discrete time series; For safety attenuation coefficient, The value ranges from 0.95 to 0.99. The basis for determining this parameter is to ensure that within 200ms to 500ms after the loss of control signal, the output energy of the hydraulic system can smoothly drop below the safety threshold (e.g., 10% of the rated output), so that the robotic arm exhibits a damped stopping characteristic, rather than a violent sudden stop or uncontrolled free movement.

[0212] To verify the effectiveness of the intelligent speed control system for automotive dismantling hydraulic equipment proposed in this invention, an experimental platform was built, comprising an edge controller 100, a servo hydraulic station, and a full-size dismantling shearing machine. Two typical working conditions were selected for comparative testing:

[0213] Condition A (High-rigidity component): Disassembling a car B-pillar (high-strength boron steel, multi-layer welded structure);

[0214] Condition B (Low-rigidity components): Disassemble the door panel and the internal sound-absorbing cotton filling.

[0215] The experiment compares the "intelligent impedance adaptive control strategy" (hereinafter referred to as "this scheme") proposed in this invention with the traditional "constant flow / pressure cutoff control strategy" (hereinafter referred to as "traditional scheme").

[0216] Impedance identification and frequency domain characteristic verification:

[0217] Please see the appendix Figure 3 The figure shows a comparison of the power spectral density (PSD) of the feedback signal from the accelerometer 23 after FFT transformation by the feature extraction module 102 under different operating conditions.

[0218] In the experiment, the excitation generation module 101 output frequency Flutter signal.

[0219] Curve A (solid line) represents the response under operating condition A (column B). It can be seen that, in A significant energy peak (-10 dB / Hz) was observed at that location, and the signal-to-noise ratio... Based on this, the impedance identification module 103 determines that the physical coupling is effective and the stiffness index is high, and the system accurately identifies high-stiffness loads.

[0220] Curve B (dashed line) represents the response under condition B (door panel). Due to the damping effect of the sound-absorbing cotton and thin plate, the signal energy at 25Hz is significantly attenuated (-45dB / Hz) and buried in the ambient noise. Based on this, the impedance identification module 103 determines that the stiffness index is low, avoiding misjudgment.

[0221] Experimental results show that the system can accurately distinguish the material properties of workpieces within 500ms through frequency domain features, with an accuracy rate of over 98%.

[0222] Adaptive speed regulation and operational stability verification:

[0223] Please see the appendix Figure 4 The figure shows the time-domain response curves of the system's key state variables during continuous shearing operations.

[0224] From top to bottom in the diagram are: equivalent stiffness index, flow control command, and shearing hydraulic cylinder pressure.

[0225] Stage T1 (0-2s): In the initial stage of contact between the shear arm and the workpiece, the stiffness index is low. The collaborative calculation module 104 outputs the full flow command (100%) to achieve rapid empty stroke closure.

[0226] Stage T2 (2-4s): As the shearing deepens and reaches the internal reinforcing ribs, the stiffness index rapidly rises to 85. This solution (solid line) responds quickly, reducing the flow command to 35% through the S-shaped mapping function, and performing strong shearing in a high-pressure, low-speed mode; while the traditional solution (gray dashed line) maintains a high flow rate, causing the system pressure to rapidly reach the overflow threshold at 3.5s, triggering high-frequency oscillations.

[0227] Performance Comparison: The pressure curve of this solution rises smoothly with no obvious pressure overshoot; compared with the traditional solution, the pressure fluctuation amplitude of the hydraulic system is reduced by 65%, effectively avoiding engine stalling.

[0228] Fracture impact and active energy absorption verification:

[0229] Please see the appendix Figure 5The figure shows a comparison of the pressure response of the hydraulic cylinder of the pressure arm at the moment of workpiece fracture, verifying the effect of active energy absorption control.

[0230] time The workpiece undergoes brittle fracture, and the load suddenly disappears.

[0231] Traditional solution (dashed line): Due to the lack of back pressure regulation, the pressure inside the hydraulic cylinder of the pressure arm drops instantly from 20MPa, and the mechanical rebound generates a reverse impact peak of up to 15MPa, with an oscillation convergence time of up to 800ms.

[0232] In this scheme (solid line): the collaborative calculation module 104 detects a sudden change in the differential shear pressure and immediately triggers an active energy-absorbing back pressure command. The pressure in the hydraulic cylinder of the pressure arm is unloaded in a controlled manner through the relief valve at the moment of fracture, the peak pressure is limited to within 5MPa, and it stabilizes rapidly within 200ms.

[0233] Conclusion: This invention reduces mechanical impact energy by approximately 70% through active variable damping control, significantly extending the service life of the pressure arm pin and hydraulic lines.

[0234] Comprehensive Energy Efficiency Data Comparison Table:

[0235] The statistical data based on 100 consecutive cyclic shear tests are shown in the table below:

[0236] Performance indicators Traditional solution Invention Solution Optimization range Mean shear period 12.5 seconds 10.8 seconds Efficiency improved by 13.6% Hydraulic oil temperature rise / hour 15.2°C 9.8°C Heat loss reduced by 35.5% Peak impact acceleration 4.5g 1.2g Mechanical impact reduced by 73.3% Fuel consumption rate 18.5L / h 15.2L / h Energy saving of 17.8%

[0237] In summary, this invention significantly improves the safety and energy efficiency of equipment while ensuring operational efficiency through frequency domain impedance identification and cross-end collaborative control.

Claims

1. An intelligent speed control system for automotive dismantling hydraulic equipment, characterized in that, include: The pressure arm assembly (20) and the shear arm assembly (30) are respectively equipped with a pressure arm hydraulic cylinder (21) and a shear arm hydraulic cylinder (31); the return oil circuit of the pressure arm hydraulic cylinder (21) is provided with a proportional relief valve (22), and the end of the pressure arm is provided with an acceleration sensor (23); the main oil inlet circuit of the shear arm hydraulic cylinder (31) is provided with an electro-hydraulic servo valve (32), and the cylinder is provided with a pressure sensor (33); the system also includes a temperature sensor (34) for collecting hydraulic oil temperature and an edge controller (100) for communicating with the above components; The edge controller (100) is configured as follows: In response to external operation commands, the valve port control command with superimposed high frequency excitation characteristics is output to the electro-hydraulic servo valve (32) to drive the shearing hydraulic cylinder (31) to generate hydraulic pulsation; Receive the time-domain vibration signal from the accelerometer (23) and the pressure signal from the pressure sensor (33), and calculate the equivalent stiffness index of the workpiece; Based on the equivalent stiffness index and time-domain vibration signal, an adaptive flow feedforward command that is negatively correlated with stiffness is output to the electro-hydraulic servo valve (32), and a dynamic back pressure setting command configured to absorb vibration energy is output to the proportional relief valve (22). The logic by which the edge controller (100) generates a dynamic back pressure setting command for the proportional relief valve (22) includes: The static reference pressure is calculated based on the equivalent stiffness index, and the static reference pressure is positively correlated with the equivalent stiffness index. The time-domain vibration signal collected by the accelerometer (23) is integrated and high-pass filtered to obtain the vibration velocity after filtering out low-frequency drift; Multiplying the vibration velocity by the virtual damping coefficient yields the dynamic damping pressure compensation value. The static reference pressure and the dynamic damping pressure compensation value are superimposed as the basic control command sent to the proportional relief valve (22).

2. The intelligent speed control system for automotive dismantling hydraulic equipment according to claim 1, characterized in that, The logic by which the edge controller (100) generates the valve port control command includes: Obtain the basic bias signal required to maintain the current contact pressure, wherein the value of the basic bias signal is set to be greater than the dead zone start voltage of the valve core of the electro-hydraulic servo valve (32); A periodic fluctuation component is superimposed on the basic bias signal, and the fluctuation component has a preset detection frequency and a preset excitation amplitude; The preset detection frequency is set to a value that is lower than the upper limit of the frequency response bandwidth of the electro-hydraulic servo valve (32) and avoids the set of inherent frequencies of the mechanical structure. The preset excitation amplitude is set to a value that establishes a dynamic pressure wave inside the shearing hydraulic cylinder (31) but does not cause macroscopic displacement of the shearing arm assembly (30).

3. The intelligent speed control system for automotive dismantling hydraulic equipment according to claim 2, characterized in that, The logic by which the edge controller (100) calculates the equivalent stiffness index of the workpiece includes: Windowing modulation and fast discrete Fourier transform are performed on the time-domain vibration signal fed back by the accelerometer (23) and the pressure signal fed back by the pressure sensor (33) to generate frequency domain power spectral density data. Extract the input pressure amplitude and output acceleration amplitude at the preset detection frequency; The effective dynamic excitation force acting on the workpiece input end is calculated based on the input pressure amplitude, and the inertial response force of the pressure arm assembly (20) is calculated based on the output acceleration amplitude. The mechanical transmission rate is obtained by calculating the ratio between the inertial response force and the effective dynamic excitation force, and the mechanical transmission rate is normalized and mapped to the equivalent stiffness index.

4. The intelligent speed control system for automotive dismantling hydraulic equipment according to claim 3, characterized in that, Before calculating the equivalent stiffness exponent, the edge controller (100) performs a physical coupling effectiveness check: Based on the frequency domain power spectral density data, the total power of the main lobe signal in the neighborhood of the preset detection frequency is calculated; Select noise reference frequency bands located on both sides of the preset detection frequency, and calculate the average noise power density within the noise reference frequency bands; The ratio of the total power of the main lobe signal to the average noise power density is calculated as the frequency domain signal-to-noise ratio index. If the frequency domain signal-to-noise ratio is greater than the preset confidence threshold, the physical coupling is determined to be effective and stiffness calculation is performed; if the frequency domain signal-to-noise ratio is less than or equal to the preset confidence threshold, the physical coupling is determined to be invalid and a default safe flow command is generated.

5. The intelligent speed control system for automotive dismantling hydraulic equipment according to claim 1, characterized in that, The logic by which the edge controller (100) generates the adaptive traffic feedforward instruction includes: Parse the operation instructions to obtain the normalized traffic demand instructions; The flow correction gain coefficient is calculated using a nonlinear attenuation model based on a sigmoid function, wherein the sigmoid function is configured to cause the flow correction gain coefficient to decrease rapidly after the equivalent stiffness index exceeds the inflection point stiffness value of the operating condition. A pressure saturation correction logic is introduced to forcibly reduce the flow output when the actual shear pressure approaches the system overflow pressure. The rated flow rate of the electro-hydraulic servo valve (32), the flow demand command, the flow correction gain coefficient, and the coefficient of the pressure saturation correction logic output are multiplied together to generate the final valve opening flow command.

6. The intelligent speed control system for automotive dismantling hydraulic equipment according to claim 1, characterized in that, The edge controller (100) is also configured with fracture impact suppression logic: Real-time monitoring of the time-delayed component of the input pressure signal of the shearing hydraulic cylinder (31); When the detected time component is lower than the preset pressure drop threshold, it is determined that the workpiece has broken. A pulse locking pressure with a preset duration is superimposed on the basic control command to increase the overflow opening threshold of the hydraulic cylinder (21) of the pressure arm to resist the rebound impact.

7. The intelligent speed control system for automotive dismantling hydraulic equipment according to claim 1, characterized in that, The edge controller (100) also executes timing synchronization logic for control commands: The hydraulic oil temperature fed back by the temperature sensor (34) and the system pressure fed back by the pressure sensor (33) are collected in real time. Based on the viscosity change caused by oil temperature and the bulk modulus change caused by pressure, the total response lag time of the hydraulic system is estimated. The phase lead step is calculated based on the ratio of the total response lag time to the control cycle period. Based on the phase advance step size, the currently calculated control parameters are preloaded into the buffer storage area, so that the effective time of the control command is aligned with the physical action time of the hydraulic system on the time axis.

8. The intelligent speed control system for automotive dismantling hydraulic equipment according to claim 3, characterized in that, When calculating the effective dynamic excitation force, the edge controller (100) is configured as follows: The theoretical driving force is calculated based on the product of the input pressure amplitude and the effective pressure-bearing area of ​​the shear hydraulic cylinder (31); Obtain the pre-calibrated equivalent friction loss force, which characterizes the dynamic seal friction and viscous damping during the movement of the hydraulic cylinder; Subtracting the equivalent friction loss force from the theoretical driving force yields the effective dynamic excitation force required for stiffness calculation.

9. The intelligent speed control system for automotive dismantling hydraulic equipment according to claim 1, characterized in that, The edge controller (100) is also configured with fault-oriented safety logic: Periodically verify the sampling voltage and physical gradient change rate of the acceleration sensor (23) and the pressure sensor (33); When the sampled voltage exceeds the threshold for determining whether the circuit is open or short-circuited, the rate of change of the physical gradient exceeds the maximum response speed allowed by the physical system, or a logical interlock conflict occurs between the flow command and the feedback pressure, a fault state is triggered. In the fault state, the control strategy based on the equivalent stiffness index is shielded, the electro-hydraulic servo valve (32) is switched to manual speed limiting mode, and the proportional relief valve (22) is locked to constant safety pressure holding mode.

Images