An automated control system for stabilizing the torque of a dialysis machine articulated arm.

Abstract

This application relates to the field of automated control technology for medical devices, and discloses an automated control system for torque stabilization of a dialysis machine articulated arm. The system includes a main control computing module, and a sensing and detection module, an environmental sensing module, and a servo drive module, all connected to the main control computing module via an industrial data bus. The sensing and detection module synchronously collects real-time torque, angular velocity, and absolute angular displacement data, while the environmental sensing module collects real-time ambient temperature data. The main control computing module receives the above data, performs calculations to construct an ambient temperature feedforward compensation coefficient, and generates motion control commands by combining the dynamic convergence index and accumulated frictional work. These commands are then sent to the servo drive module to drive the dialysis machine articulated arm to perform unidirectional continuous rotational motion or small angular displacement reversing motion. This invention decouples the interference of break-in heat generation and external environmental temperature difference on the viscosity of damping grease, eliminates false convergence states, improves the accuracy of torque determination, and reduces the incidence of jamming during assembly.

Description

Technical Field

[0001] This invention relates to the field of automated control technology for medical devices, specifically an automated control system for stabilizing the torque of the articular arm of a dialysis machine. Background Technology

[0002] During the manufacturing process of dialysis machines, the damping torque of the articulated arm needs to be controlled within a set range to ensure that medical staff have a suitable feel and stable hovering positioning when pushing and pulling. Typically, the articulated arm needs to undergo a rotational break-in test at the assembly station. Existing automated control systems mostly rely on motors to drive the articulated arm to rotate continuously and read the values ​​of torque sensors in real time. When the torque amplitude is detected to enter the design tolerance zone, the assembly break-in is deemed qualified and the operation is terminated.

[0003] However, the aforementioned conventional control methods have limitations in actual manufacturing. During the initial break-in period of the articulated arm, the microscopic protrusions on the internal friction surfaces undergo shearing and plastic rheology, converting mechanical friction work into heat energy that accumulates locally. The dynamic viscosity of the damping grease decreases with increasing temperature, resulting in a reduction in the macroscopic torque amplitude and a stable state. Existing systems rely solely on the torque amplitude as a shutdown criterion, failing to recognize this steady-state illusion caused by thermal effects. This leads to a rebound and shift in the actual damping torque after the workpiece is removed from the test environment and cooled.

[0004] Because damping grease is sensitive to ambient temperature, temperature differences in industrial workshops caused by day-night cycles or seasonal changes can alter its initial viscosity. Existing systems do not incorporate macroscopic environmental boundary conditions as feedforward variables into the control closed loop, relying solely on the a posteriori state generated by mechanical friction. This leads to premature shutdowns at low temperatures due to the initial viscosity of the damping grease, and excessive cutting intervention at high temperatures. Under continuous unidirectional rotation, damping grease can accumulate in the edge gaps due to mechanical compression, causing localized dry friction and jamming risks at the core contact surface. Conventional testing systems lack dynamic intervention mechanisms for fluid boundary lubrication. Regarding data acquisition and communication architecture, existing systems are prone to transmission link delays when processing data from multi-source heterogeneous sensors. The data from the analog strain conversion link and the digital photoelectric encoder link have a time phase difference, failing to ensure strict alignment of torque and angular velocity characteristics on the physical time axis, thus affecting the accuracy of upper-level computing platforms in reconstructing features such as work integrals. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention provides an automated control system for stabilizing the torque of a dialysis machine articulated arm. This system solves the problems of existing technologies that rely solely on a single torque amplitude as a shutdown criterion, fail to decouple the interference of frictional heat effects and ambient temperature changes, and lack an effective lubrication intervention mechanism and a multi-source data synchronization architecture, resulting in insufficient accuracy and stability in determining the torque of the articulated arm.

[0006] To achieve the above objectives, the present invention is implemented through the following technical solution: an automated control system for stabilizing the torque of a dialysis machine articulated arm, comprising a main control computing module internally equipped with a data processing unit and a logic judgment unit, and a sensing and detection module, an environmental perception module, and a servo drive module connected to the main control computing module via an industrial data bus;

[0007] The sensing and detection module includes a torque acquisition unit that synchronously acquires real-time torque data and an angular velocity acquisition unit that acquires real-time angular velocity data and absolute angular displacement data; the environmental sensing module includes a temperature acquisition unit that acquires real-time ambient temperature data; the servo drive module includes an instruction parsing unit and a drive execution unit with a servo motor, a reduction mechanism, and a spindle fixture.

[0008] The main control calculation module receives the real-time torque data, real-time angular velocity data, absolute angular displacement data, and real-time ambient temperature data. It constructs an ambient temperature feedforward compensation coefficient through the data processing unit, generates motion control commands through the logic judgment unit, and sends them to the command parsing unit. The drive execution unit drives the main spindle fixture tooling, which in turn drives the dialysis machine articulated arm to perform unidirectional continuous rotation or small angular displacement reversal motion.

[0009] As a further embodiment of the present invention, the data processing unit is used to set a sliding data window, perform time-domain statistics on the real-time torque data within the sliding data window, and calculate the mean dynamic torque and the standard deviation of dynamic torque within the current sliding data window.

[0010] As a further aspect of the present invention, the data processing unit is used to perform multi-dimensional matrix mapping addressing in a preset internal parameter mapping table based on the pre-extracted initial static friction torque peak value, and to match and obtain low-frequency trend weighting coefficients, high-frequency fluctuation weighting coefficients, and thermal attenuation coefficients; the data processing unit is used to combine the low-frequency trend weighting coefficients, the high-frequency fluctuation weighting coefficients, the absolute value of the first derivative of the dynamic torque mean with respect to time, and the dynamic torque standard deviation to calculate and obtain a dynamic convergence index; the dynamic convergence index is calculated by adding the product of the low-frequency trend weighting coefficients and the absolute value of the first derivative of the dynamic torque mean with respect to time, to the product of the high-frequency fluctuation weighting coefficients and the dynamic torque standard deviation.

[0011] As a further aspect of the present invention, the data processing unit is used to perform continuous time integration on the real-time torque data and the real-time angular velocity data to obtain the cumulative frictional work.

[0012] As a further embodiment of the present invention, the temperature acquisition unit is used to smooth the real-time ambient temperature data and output a smoothed real-time ambient temperature, which is received by the main control calculation module; the data processing unit is used to calculate and construct the ambient temperature feedforward compensation coefficient by combining the pre-calibrated damping grease viscosity-temperature sensitivity coefficient, the smoothed real-time ambient temperature, and the pre-calibrated standard reference ambient temperature; the logic determination unit is used to extract the pre-stored basic static convergence threshold with a value range of 0.01 to 0.10, and calculate the global coupling dynamic convergence determination threshold in real time using the thermal decay coefficient, the cumulative friction work, the ambient temperature feedforward compensation coefficient, and the basic static convergence threshold; the global coupling dynamic convergence determination threshold is calculated by multiplying the ambient temperature feedforward compensation coefficient, the basic static convergence threshold, and an exponential function with the natural constant as the base and the negative value of the product of the thermal decay coefficient and the cumulative friction work as the exponent.

[0013] As a further aspect of the present invention, the data processing unit is used to set a compliance upper limit threshold of 1.5 and a compliance lower limit threshold of 0.5. When the ambient temperature feedforward compensation coefficient deviates from the safe range formed by the compliance lower limit threshold and the compliance upper limit threshold, the logic determination unit is used to activate the edge saturation pinning drive to rigidly limit the ambient temperature feedforward compensation coefficient to the nearest boundary extreme value.

[0014] As a further aspect of the present invention, when the average dynamic torque value does not enter the set target tolerance zone and the dynamic convergence index is greater than the intervention upper limit threshold set to 3 to 5 times the basic static convergence threshold, the logic determination unit is used to generate a speed reduction protection command and send it to the servo drive module to reduce the angular velocity of the servo motor according to a set slope; when the average dynamic torque value is within the target tolerance zone and the time derivative of the standard deviation of the dynamic torque exceeds the preset trigger limit, the logic determination unit is used to generate a lubricant redistribution control command to instruct the servo drive module to stop the unidirectional continuous rotational motion and switch to asymmetric micro-stroke reciprocating commutation motion.

[0015] As a further aspect of the present invention, when the average dynamic torque is within the target tolerance zone and the dynamic convergence index is less than the global coupling dynamic convergence judgment threshold and the duration of this condition reaches the set judgment period, the logic judgment unit is used to issue a process completion signal, instructing the servo drive module to execute the set maximum braking torque stop action allowed by the system and lock the current workpiece assembly state.

[0016] As a further embodiment of the present invention, the drive execution unit is used to respond to the diagnostic frequency sweep command issued by the main control computing module, control the servo motor to drive the spindle fixture tooling, perform a single micro angular displacement detection motion, and cooperate with the sensing detection module to capture the initial static friction torque peak value; the drive execution unit is used to respond to the lubricant redistribution control command, switch the servo motor to an asymmetric micro-stroke reciprocating commutation motion mode, and call the internally loaded preset forward commutation angular acceleration and reverse commutation angular acceleration to execute the asymmetric micro-stroke reciprocating commutation motion.

[0017] As a further embodiment of the present invention, the torque acquisition unit is equipped with a dynamic torque sensor; the input end of the dynamic torque sensor is connected to the output shaft of the reduction mechanism through a high-rigidity diaphragm coupling, and the output end of the dynamic torque sensor is connected to the spindle fixture tooling; the sensing and detection module is used to utilize the distributed clock synchronization mechanism of the industrial data bus to synchronize and package the torque data channel and angular velocity data channel and attach the same clock timestamp.

[0018] This invention provides an automated control system for stabilizing the torque of the articulated arm of a dialysis machine. It has the following beneficial effects:

[0019] 1. This invention obtains the dynamic convergence index by combining the first derivative of the mean dynamic torque and the standard deviation of the dynamic torque, and obtains the cumulative friction work by continuously integrating the real-time torque and angular velocity. It also constructs the ambient temperature feedforward compensation coefficient by combining the smoothed ambient temperature data, and finally calculates the global coupling dynamic convergence judgment threshold. The above-mentioned control logic based on multi-dimensional parameter coupling decouples the false convergence state caused by the decrease in damping grease viscosity due to the heat generated during break-in, and eliminates the interference of the workshop ambient temperature difference on the initial judgment boundary of the damping grease, thereby improving the accuracy of the joint arm torque judgment.

[0020] 2. When the control system of the present invention detects that the average dynamic torque is within the target range and the time derivative of the standard deviation of the dynamic torque exceeds the trigger limit, it will control the servo motor to stop unidirectional continuous rotation and switch to asymmetric micro-stroke reciprocating commutation motion. This structure uses the set commutation angle acceleration to destroy the shear yield grid of the damping grease in the oil-deficient area, and uses the asymmetric reciprocating displacement to build a fluid pressure gradient, guiding the damping grease to flow back and penetrate into the core friction surface, thereby reducing the jamming rate during the assembly and break-in process.

[0021] 3. In terms of hardware structure and data transmission, this invention connects the input end of the dynamic torque sensor to the output shaft of the reduction mechanism through a high-rigidity diaphragm coupling, and uses the distributed clock synchronization mechanism of the industrial data bus to add the same clock timestamp to the torque and angular velocity data channels. This eliminates the hysteresis and deformation error of traditional flexible transmission in the mechanical link, and eliminates the clock phase difference when heterogeneous data is transmitted across channels in the electrical link, thus providing a reliable data foundation for system operation and control. Attached Figure Description

[0022] Figure 1 This is a block diagram of the overall architecture of the automated control system for stabilizing the torque of the articulated arm of a dialysis machine provided in an embodiment of the present invention;

[0023] Figure 2 This is a flowchart illustrating the overall steps of the automated control method for stabilizing the torque of the articulated arm of a dialysis machine provided in an embodiment of the present invention.

[0024] Figure 3 The flowchart shows the specific operating logic of the adaptive convergence determination algorithm inside the main control computing module.

[0025] Figure 4 This is a flowchart illustrating the specific mechanical control sequence and physical execution logic of the servo drive module.

[0026] Figure 5 This is a diagram showing the evolution trajectory of process convergence determination under low-temperature conditions according to the present invention.

[0027] Figure 6 This is a comparison chart showing the long-term stability of the joint arm damping torque after the break-in process of this invention is completed. Detailed Implementation

[0028] 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.

[0029] Please see the appendix Figure 1 This invention provides an automated control system for stabilizing the torque of a dialysis machine articulated arm, comprising: a main control computing module, a servo drive module, a sensor detection module, and an environmental perception module.

[0030] The main control computing module establishes communication connections with the servo drive module, sensor detection module, and environmental perception module via an industrial data bus. Internally, the main control computing module includes a data processing unit and a logic decision unit. The data processing unit receives the underlying physical data returned by each sensor node and performs discretized mathematical operations. The logic decision unit generates and issues motion control commands based on the operation results.

[0031] The sensing and detection module is located at the assembly station and specifically includes a torque acquisition unit and an angular velocity acquisition unit. The torque acquisition unit houses a dynamic torque sensor, mechanically mounted coaxially in series with the test spindle, used to synchronously acquire real-time torque data of the dialysis machine's articulated arm during rotational break-in. The angular velocity acquisition unit houses an absolute encoder, integrated into the feedback end of the servo motor, used to acquire real-time angular velocity and absolute angular displacement data from the servo drive module.

[0032] The environmental sensing module includes a temperature acquisition unit, which is deployed on the periphery of the test station where the control system is located to collect real-time environmental temperature data. The main control computing module receives the real-time environmental temperature data and constructs an environmental temperature feedforward compensation coefficient through the data processing unit.

[0033] The servo drive module includes an instruction parsing unit and a drive execution unit. The instruction parsing unit receives position and speed instructions output by the main control calculation module. The drive execution unit includes a servo motor, a reduction mechanism, and a spindle fixture. It drives the spindle fixture using signals from the instruction parsing unit, thereby causing the dialysis machine's articulated arm to perform unidirectional continuous rotational motion or small angular displacement reversing motion.

[0034] Please see the appendix Figure 2 This invention provides an automated control method for stabilizing the torque of a dialysis machine articulated arm, comprising the following steps:

[0035] S10, the sensing and detection module synchronously collects real-time torque data of the dialysis machine articulated arm during the rotation break-in process through the torque acquisition unit, and collects real-time angular velocity data and absolute angular displacement data of the servo drive module through the angular velocity acquisition unit;

[0036] S20, the environmental sensing module collects real-time environmental temperature data from the surrounding environment of the test station through the temperature acquisition unit;

[0037] S30, the main control computing module receives the underlying physical data and real-time ambient temperature data returned by each sensor node, performs discretized mathematical operations through the data processing unit, and constructs the ambient temperature feedforward compensation coefficient.

[0038] S40, the main control computing module generates and issues motion control commands based on the calculation results through the logic judgment unit;

[0039] S50, the instruction parsing unit of the servo drive module receives position and speed instructions, and drives the spindle fixture tooling through the drive execution unit to drive the articulated arm of the dialysis machine to perform unidirectional continuous rotational motion or small angular displacement reversal motion.

[0040] The general technical principle upon which this embodiment relies lies in the fluid boundary lubrication transition mechanism and the law of conservation of energy in tribology. During the initial rotational break-in period of the dialysis machine's articulated arm, the microscopic protrusions on the internal friction pair surfaces undergo intense shearing and plastic rheology. Accompanying this is the accumulation of heat energy converted from frictional work, causing a localized instantaneous temperature increase in the damping grease, which leads to a nonlinear decrease in its dynamic viscosity. This manifests macroscopically as a false reduction and stabilization of the torque amplitude. To accurately eliminate this steady-state illusion caused by thermal effects, it is necessary to couple spatial, temporal, and thermodynamic energy scales in multiple dimensions. The main control computing module, as the core computing platform of the control system, primarily relies on its built-in data processing unit and logic decision unit to execute physical feature reconstruction and adaptive control strategies. The data processing unit and logic decision unit within the main control computing module can be implemented using lower-level hardware structures such as an industrial control computer with a real-time operating system, a central processing unit of a programmable logic controller, or a digital signal processor; no specific limitations are imposed here.

[0041] For the low-level programming implementation of matrix mapping addressing and calculus operations in industrial control hardware, those skilled in the art can use general structured text languages ​​or ladder diagrams. The code compilation and execution mechanism is a well-known technology in this field and will not be elaborated here.

[0042] The specific operating logic of the main control computing module is implemented through the following steps:

[0043] S100, the data processing unit reconstructs the high-frequency acquired real-time torque data based on a set sliding time window. To achieve strict alignment of multi-source data in terms of time and operating conditions, the main control calculation module periodically and synchronously receives real-time dynamic torque data acquired by the sensing module and processed by low-pass filtering for noise reduction. With real-time angular velocity In this embodiment, the data processing unit sets a sliding data window with a time length greater than zero. By using this sliding data window Real-time dynamic torque within Perform time-domain statistics to calculate the average dynamic torque within the current window. With dynamic torque standard deviation During this process, the data window slides. The length is preferably in the range of 0.5s to 2.0s, and its determination is based on the time period required for the articulated arm to rotate one revolution, so as to ensure that the window contains complete spatial rotation information.

[0044] S200, the data processing unit performs multi-dimensional parameter addressing and calculates the dynamic convergence exponent and cumulative frictional work. This is based on the initial static friction torque peak value extracted from the diagnostic frequency sweep action in the preceding process. The data processing unit uses this as a unique index key and performs multidimensional matrix mapping addressing in a pre-defined internal parameter mapping table. If singular boundary values ​​are encountered during the addressing process, a nearby node linear interpolation method is automatically used for smoothing, thereby matching and obtaining a set of control parameters suitable for the current specific workpiece tolerance state, specifically including low-frequency trend weighting coefficients. High-frequency fluctuation weighting coefficient and thermal attenuation coefficient Among them, the low-frequency trend weighting coefficient The value range is from 0.1 to 2.0, which is the high-frequency fluctuation weighting coefficient. The value ranges from 0.5 to 5.0, and the determination of both is based on the weighted allocation of the physical correlation between the microscopic peak cutting stage and the macroscopic damping stabilization stage. The data processing unit calculates the dynamic convergence index by combining the derivative and standard deviation of the mean characteristics. The calculation formula is:

[0045] ;

[0046] In the formula, The dynamic convergence exponent; The low-frequency trend weighting coefficient; For high-frequency fluctuation weighting coefficients; The standard deviation of dynamic torque; The absolute value of the first derivative of the mean dynamic torque with respect to time is given. Its discretized form within the digital processing unit is the difference between the mean dynamic torque of the current sampling period and the mean dynamic torque of the previous sampling period, divided by the sampling period. And sampling period The formula is strictly greater than zero to avoid computational anomalies caused by the denominator approaching zero. The physical meaning of this formula lies in its comprehensive quantification of the convergence rate and dispersion of the microscopic friction state of the friction pair surface towards a macroscopic physical steady state through a weighted fusion of low-frequency trend change rate and high-frequency fluctuation rate.

[0047] To broaden the protection scope of the underlying method logic and avoid overly restrictive limitations by a single formula, the data processing unit performs real-time dynamic torque analysis. With real-time angular velocity By performing continuous-time integration, the cumulative frictional work done to quantify the frictional heat effect is obtained. The cumulative work done by friction The numerical value is expressed as real-time dynamic torque. With real-time angular velocity The product of time zero to the current time The definite integral result over the continuous time domain. The physical meaning of the above integration process is that, based on the law of conservation of energy, the work consumed by mechanical motion is converted into the cumulative thermal load of the friction interface, providing quantitative data support for subsequent thermodynamic property decoupling.

[0048] S300, the logic determination unit constructs a dynamic convergence determination threshold that includes environmental feedforward compensation and work attenuation. In this embodiment, to eliminate the nonlinear interference of workshop ambient temperature difference on the initial viscosity of damping grease, the logic determination unit receives the ambient temperature feedforward compensation coefficient transmitted by the environmental sensing module. And extract the pre-stored basic static convergence threshold. Basic static convergence threshold The value range is set to 0.01 to 0.10, and its determination is based on the ideal convergence index of a fully run-in qualified sample under standard temperature conditions. The logic decision unit utilizes the thermal decay coefficient. Work done by accumulated friction Real-time calculation of the global coupling dynamic convergence judgment threshold The calculation formula is:

[0049] ;

[0050] In the formula, The threshold for determining the dynamic convergence of the global coupling; This is the feedforward compensation coefficient for ambient temperature; The basic static convergence threshold; Represents an exponential function with the natural constant as its base; thermal decay coefficient. The value range is 1×10 -6 J -1 Up to 1×10 -4 J -1 The basis for its determination is the viscosity-temperature decay curve characteristics of the damping grease at a specific shear rate. This is the amount of work done by accumulated friction. The physical meaning of this formula is to construct a dynamic decision boundary that actively tightens as physical work is consumed. This forces the system to meet more stringent microscopic stability requirements before terminating the process under conditions of increased accumulated heat load and thinning of damping grease. This completely decouples the algorithm from the underlying layer and counteracts the false convergence state caused by thermal effects.

[0051] In S400, the logic decision unit triggers process intervention commands based on the evolution trajectory of the dynamic convergence exponent and the average torque value. To avoid the bias and false triggering caused by a single physical extreme value determination, the logic decision unit uses a composite weighted logic based on a multi-dimensional two-dimensional phase trajectory for control intervention. When the dynamic torque average value... Not within the target tolerance zone And the dynamic convergence index When the speed exceeds the set intervention upper limit threshold, the logic judgment unit generates a speed reduction protection command and sends it to the servo drive module, reducing the angular velocity of the servo motor at a set slope to suppress abnormal heat generation at the friction interface. The target tolerance zone... The value range is determined based on the specifications and feel design of the dialysis machine's articulated arm, with the upper limit threshold set as the basic static convergence threshold. 3 to 5 times.

[0052] Based on the above condition monitoring, when the average dynamic torque Within the target tolerance zone Within the range, but the standard deviation of dynamic torque When the time derivative exceeds the preset trigger limit, the logic determination unit generates a lubricant redistribution control command, instructing the servo drive module to stop unidirectional continuous rotation and switch to asymmetric micro-stroke reciprocating motion, thereby inducing a hydrodynamic pumping effect. The trigger limit is determined based on the instantaneous abrupt change rate of high-frequency roughness noise within the sliding window, which accurately characterizes the physical state where the damping grease is completely squeezed out in a local micro-region, leading to dry friction jamming at the contact surface.

[0053] The S500 logic decision unit performs dual logic gate state latching and outputs the final process signal. The logic decision unit has a dedicated logic core based on time dwell parameters. This is done when the control system simultaneously meets the static tolerance compliance conditions. and dynamic anti-counterfeiting convergence conditions Furthermore, the duration of both conditions must reach the set judgment period. At that time, the logic determination unit confirms that the system has reached a physical steady state. In this embodiment, the determination period is... The value range is set to 3s to 10s, and it is determined based on ensuring that the damping torque drift does not occur after the workpiece has undergone at least two complete rotation cycles after it has been removed from the micro-cutting process. At this time, the logic determination unit issues a process completion signal, instructing the servo drive module to perform the maximum braking torque stop action and lock the current workpiece assembly state.

[0054] Regarding system security boundaries, the monitoring mechanism has internally set maximum work limits. Maximum work limit The value range is set based on the maximum thermal distortion limit that the workpiece's mechanical structure can withstand. The calculated cumulative frictional work... Greater than the maximum work limit Furthermore, since the system did not trigger the process completion signal, the logic determination unit interrupted the break-in main program, threw an abnormal scrap code to the external bus, and forcibly executed a system shutdown. This multi-dimensional composite determination logic ensures that the output process result depends not only on the current amplitude state, but also on its deep physical stability along the thermodynamic evolution path.

[0055] See attached document Figure 3 The specific mechanical control sequence involved in this embodiment is based on the general technical principle of electromechanical energy conversion and the pumping effect in non-Newtonian fluid dynamics. The digital characteristic judgment and intervention logic output by the control system needs to cross the boundary between digital communication and physical execution, and be transformed into the actual electromagnetic torque and spatial angular displacement of the servo spindle. As the physical execution mechanism of the system, the servo drive module mainly relies on the built-in instruction parsing unit and drive execution unit to transform the logical instructions into precise mechanical control actions. For the underlying space vector pulse width modulation and position, speed, and current three-closed-loop feedback control algorithm of the servo motor, those skilled in the art can use mature commercial industrial Ethernet servo driver firmware to implement it. Its signal hard-wired connection and drive mechanism are well-known technologies in the field and will not be described in detail here.

[0056] The specific operating logic of the servo drive module is implemented through the following steps:

[0057] S600, the instruction parsing unit performs the underlying protocol conversion from logic control instructions to mechanical motion instructions. To ensure strict alignment of multi-source data in terms of time and operating conditions, the instruction parsing unit periodically receives a set of logic instructions from the main control computing module via the industrial data bus. This set of logic instructions includes speed reduction protection instructions, lubricant redistribution sequence instructions, and shutdown instructions. In this embodiment, as an optional implementation, the bus communication cycle is set to 1ms to 4ms. Based on the current servo absolute position encoding data, the instruction parsing unit calculates the aforementioned discrete logic judgment instructions in real time into a target angular velocity that can be directly executed by the lower-level driver. Target angular acceleration and target angular displacement The specific values ​​of the aforementioned target motion parameters are dynamically generated by the main control calculation module in a linear proportion based on the current degree of deviation from the target tolerance zone. Based on this continuous instruction parsing mechanism, the system establishes a one-to-one mapping coupling channel between the mathematical feature evolution at the algorithm level and the mechanical actions at the physical level, thus providing a low-latency communication foundation for subsequent adaptive targeting process intervention.

[0058] In the S700, the drive actuator, in conjunction with the pre-diagnostic process, induces the initial static friction torque peak. Before the main break-in program starts, the drive actuator responds to the diagnostic frequency sweep command, controlling the servo motor at the set diagnostic angular velocity. The drive spindle fixture performs a single, minute angular displacement detection motion. In this embodiment, the angular velocity is diagnosed. The value ranges from 0.5° / s to 2.0° / s, allowing for the detection of minute angular displacements. The extreme values ​​are controlled within 3°. The above parameters are determined to avoid irreversible macroscopic plastic deformation of the internal mechanical structure of the dialysis machine's articulated arm. The technical purpose of performing this low-speed micro-stroke action is to ensure that the micro-friction pairs of the dialysis machine's articulated arm just cross the critical shear yield point from rest to sliding through low-energy power input, while limiting angular displacement to prevent it from prematurely entering the macroscopic micro-peak cutting stage. This, in conjunction with the sensing module, captures the initial static friction torque peak. .

[0059] The S800 actuator responds to lubricant redistribution sequence commands by performing fluid intervention actions. When the control system detects a risk of localized dry friction and issues a lubricant redistribution sequence command, the actuator disconnects the current unidirectional continuous rotation output and switches the servo motor to an asymmetric micro-stroke reciprocating commutation motion mode within one communication cycle of the internal bus. The actuator internally loads and executes preset high-dynamic gait parameters, including a set positive commutation angular acceleration. With reverse commutation angular acceleration The absolute values ​​are all within the range of 500° / s 2 Up to 2000° / s 2 The determination of this range is based on overcoming the static starting resistance of high-viscosity damping grease in a lack of oil. The ratio of the positive to the negative angular displacement within a single reciprocating cycle is set to 1.2 to 2.0, thus creating a clear asymmetric displacement difference in space. The high angular acceleration impact can disrupt the shear yield grid of the damping grease in the lack-oil micro-region, while the asymmetric reciprocating directional displacement constructs a directional fluid pressure gradient in the narrow fit gap. This pressure gradient forces the damping grease, which was originally accumulated at the tolerance edge due to mechanical centrifugal and squeezing effects, to flow back and penetrate into the core friction surface through the fluid pumping effect. After the high-frequency dispersion of the friction interface returns to stability, the drive unit automatically releases the directional mode and resumes unidirectional continuous rotation.

[0060] The S900's drive actuator performs maximum braking torque shutdown and workpiece state latching based on safety boundaries. When the main control calculation module outputs a process completion signal or throws an exception code due to excessive accumulated friction work, the drive actuator takes over the underlying enable control of the servo motor. The drive actuator injects reverse excitation current into the stator coil of the servo motor to achieve the maximum braking torque allowed by the system. Force the rotating spindle to brake for a set time. Internal stationary. In this embodiment, the maximum braking torque... The braking time is determined based on 2.0 to 2.5 times the rated torque of the servo motor. Less than 0.1s. This rapid stop action using a brakeless ramp aims to eliminate the additional frictional heat generated during the conventional stop inertial coasting phase, ensuring that the thermodynamic state at the process determination moment is consistent with the physical state of the workpiece at its final rest on the time axis. After the spindle comes to a stop, the drive actuator maintains the servo motor's position loop in a high-gain enabled state, thereby locking the assembly angle and internal stress state of the dialysis machine's articulated arm until the external loading / unloading mechanism sends a disengagement signal.

[0061] See attached document Figure 4 In this embodiment, the general technical principle upon which the sensing and detection module relies lies in the high-fidelity conversion of electromechanical signals and anti-aliasing sampling theory. To ensure that the main control computing module is not interfered with by underlying physical noise when performing dynamic convergence feature reconstruction and work integration calculations, the sensing and detection module needs to eliminate transmission backlash in the mechanical link and suppress high-frequency aliasing in the electrical link. For the electromagnetic shielding wiring of the sensor and the underlying register settings of the analog-to-digital converter chip, those skilled in the art can use industry-standard shielded twisted-pair cables and commercial data acquisition cards to implement these features. The underlying hardware drivers and electrical isolation specifications are well-known technologies in the field and will not be elaborated upon here.

[0062] The specific operating logic of the sensing and detection module is implemented through the following steps:

[0063] S1000, the torque acquisition unit performs high-rigidity coaxial physical quantity conversion to obtain raw torque feedback. The torque acquisition unit incorporates a dynamic torque sensor, which is mechanically mounted coaxially in series with the test spindle. In this embodiment, the input end of the dynamic torque sensor is fixedly connected to the output shaft of the servo motor reduction mechanism via a backlash-free high-rigidity diaphragm coupling, and its output end is directly rigidly connected to the flange face of the spindle fixture. This direct-drive mechanical structure aims to eliminate the mechanical hysteresis and elastic deformation errors caused by traditional flexible transmissions such as belts or gears. This structure ensures that the minute damping torque changes generated by the dialysis machine's articulated arm during the rotational break-in phase can be transmitted to the sensor's elastic body in real time without phase shift or amplitude attenuation, thereby accurately converting macroscopic mechanical resistance into electrical micro-strain signals and constructing a true physical feedback source for the entire control system.

[0064] S1100, the torque acquisition unit performs high-frequency synchronous sampling and anti-aliasing filtering noise reduction. The torque acquisition unit internally incorporates a high-precision signal conditioning circuit that periodically performs analog-to-digital conversion and sampling on the analog strain signal output by the dynamic torque sensor. In this embodiment, the synchronous sampling rate... The optimal range for the synchronous sampling rate is 1000Hz to 5000Hz. This rate is determined based on the Nyquist sampling theorem to ensure sufficient data density to fully capture high-frequency transient torque fluctuations induced by grease film rupture or microscopic peak cutting. The acquired raw digital sequence contains interference from the workshop power grid and high-frequency radiation noise from the lower-level servo motor's pulse width modulation. The torque acquisition unit smooths the raw sequence using its integrated low-pass filtering algorithm, outputting real-time dynamic torque for direct use by the upper-level logic. The cutoff frequency of a low-pass filter. The frequency range is set to 100Hz to 300Hz, which is determined by the boundary between the upper limit of the workpiece friction characteristic frequency and the lower limit of the ambient high-frequency noise frequency, so as to achieve an engineering balance between preserving effective tribological dynamic characteristics and filtering out electromagnetic noise.

[0065] The S1200 angular velocity acquisition unit performs high-precision positioning acquisition of absolute spatial displacement and dynamic angular velocity. The angular velocity acquisition unit incorporates an absolute encoder, which is directly integrated into the stator feedback of the servo motor. To meet the position closed-loop control requirements of the drive module for micro-stroke reciprocating commutation movements during lubricant redistribution sequences, the absolute encoder resolution is set to no less than 20 bits, meaning it can output over one million discrete digital pulses per revolution of the servo spindle. This resolution setting is determined to ensure that minute angular displacements detected in the pre-diagnostic process can be precisely segmented, avoiding missed diagnoses due to excessively large step sizes causing static friction torque peaks to occur between data sampling points. The absolute encoder broadcasts the current spindle absolute angular displacement data to the bus in real time via an industrial high-speed serial communication protocol. Meanwhile, the firmware inside the angular velocity acquisition unit processes the absolute angular displacement data. Perform first-order discrete difference calculation based on the sampling time period to extract the real-time angular velocity of the current servo actuator. .

[0066] When performing first-order discrete difference operations, its discrete time step is... The encoder is preset to a constant positive value that is strictly greater than zero, thus preventing division-by-zero overflow anomalies caused by the time step approaching zero in the underlying algorithm design. The core technical advantage of using a physical absolute encoder instead of an incremental encoder is that even if the test station experiences an unexpected power outage after the control system performs an abnormal interruption or completes workpiece latching and shutdown, the current spatial friction phase information and internal stress residence state of the dialysis machine articulated arm will not be lost, thereby ensuring the continuity of assembly test data throughout its entire lifecycle.

[0067] S1300, the sensing and detection module performs hardware-level timestamp synchronization and alignment of multi-source heterogeneous data. Based on the aforementioned system algorithm, it is known that the main control calculation module needs to calculate the real-time dynamic torque when calculating the cumulative friction work. With real-time angular velocity The product integral is performed in the continuous time domain. Since these two sets of physical quantities originate from different hardware sensing links—the analog strain conversion link and the digital photoelectric encoder link—direct communication transmission can easily lead to time phase differences. Based on this operating condition, in this embodiment, the sensing module utilizes the distributed clock synchronization mechanism of the industrial data bus to synchronize and package the two data channels at the underlying layer. Specifically, at the instant the acquisition action is triggered, the torque acquisition unit and the angular velocity acquisition unit are forcibly stamped with the same clock timestamp by the hardware bus. The clock timestamps are uniformly distributed by the main clock source of the industrial data bus, and the maximum allowable clock phase deviation between the two data channels is strictly limited to within 1μs. This fundamentally eliminates phase misalignment caused by communication link delays, ensuring that the torque and velocity characteristics at the same physical instant are tightly aligned in digital space.

[0068] See attached document Figure 6 The general technical principle relied upon in this embodiment lies in the viscosity-temperature sensitivity characteristic of polymer rheology. Because the damping grease filling the articulated arm of a dialysis machine exhibits a high dependence on the ambient temperature, temperature differences in industrial workshops due to day-night or seasonal changes significantly alter the initial dynamic viscosity of the damping grease. If the boundary conditions of the macroscopic environment are not introduced as feedforward variables into the control closed loop in the control logic, and convergence is solely determined by the a posteriori temperature rise generated by mechanical friction, it is highly likely that the control system will experience premature shutdown and false convergence in low-temperature environments due to the initial viscosity of the damping grease, or excessive cutting intervention in high-temperature environments due to the initial thinning. To decouple this external interference at the software level, this embodiment achieves high-fidelity control across the entire temperature range by precisely constructing a physical feedforward channel. For the analysis of the underlying communication bus protocol and the selection of power supply isolation for the industrial temperature sensor in the environmental sensing module, those skilled in the art can use common platinum resistance sensors and industrial standard transmitters. The signal hard-wired connection and shielding specifications are well-known technologies in this field and will not be elaborated upon here.

[0069] The specific operational logic of the environment perception module is implemented through the following steps:

[0070] S1400, the temperature acquisition unit performs physical spatial layout and signal isolation to resist thermal radiation interference. The temperature acquisition unit includes a high-precision industrial temperature sensor. In this embodiment, the industrial temperature sensor is physically deployed on the periphery of the test station where the control system is located, and maintains a straight-line distance of 0.5m to 1.5m from the metal casing of the servo motor and the center coordinates of the dialysis machine's articulated arm. Based on heat transfer theory, the boundary of this specific spatial range is chosen so that it can accurately capture and reflect the local micro-environmental thermal background of the test workpiece, and effectively avoid the stator coil thermal radiation generated by the servo drive module during full-load operation, thereby blocking the local thermal interference caused by the intense micro-peak cutting heat generated in the initial stage of the friction pair on the environmental reference node. The probe housing of the temperature sensor is equipped with a non-metallic convection buffer cover to suppress the transient airflow temperature fluctuation interference caused by the workshop's fresh air system or personnel movement, ensuring that the physical quantities acquired by the hardware source have macroscopic thermodynamic stability.

[0071] S1500, the temperature acquisition unit performs low-frequency sampling and time-domain smoothing of the ambient temperature signal. Since the workshop ambient temperature is a slowly changing low-frequency thermodynamic parameter, to avoid introducing unnecessary high-frequency electronic noise, the temperature acquisition unit uses a low-speed sampling link to acquire real-time ambient temperature data. In this embodiment, the synchronous sampling frequency of the temperature acquisition unit is preset to 1Hz to 5Hz. Based on the time and operating condition alignment logic of multi-source data, although the motion control loop inside the main control computing module operates at a high-frequency cycle in milliseconds, due to the extremely high thermal inertia of the macroscopic ambient temperature, this low-frequency sampling combined with asynchronous data buffer reading not only ensures real-time alignment of thermodynamic boundary conditions but also effectively reduces the computational load on the central processing unit of the main control computing module. The acquired raw temperature discrete sequence is directly input to the moving average filtering algorithm module in the firmware layer for smoothing. The time window width of the moving average filter is set to 10s to 30s to further filter out discontinuous numerical jumps caused by power supply ripple, outputting a smoothed real-time ambient temperature for direct use by the upper-level main control computing module. .

[0072] The S1600's main control computing module constructs an ambient temperature feedforward compensation coefficient based on the smoothed real-time ambient temperature. The main control computing module periodically reads the smoothed real-time ambient temperature through the data processing unit. It also extracts the standard reference ambient temperature pre-calibrated in the underlying non-volatile memory. With damping grease viscosity temperature sensitivity coefficient To avoid limiting the protection range due to simply listing derived formulas, the ambient temperature feedforward compensation coefficient is used. The determination can be based on a linear thermal compensation model. Its solution logic is expressed as: damping grease viscosity temperature sensitivity coefficient. With current smooth real-time ambient temperature and standard reference ambient temperature The product of the differences is then added to a constant 1 to extract the ambient temperature feedforward compensation coefficient. In this embodiment, the standard reference ambient temperature... The value is fixed at 25℃, and its determination is based on the standard laboratory thermodynamic reference point of the damping grease supplier's rheological test. Damping grease viscosity-temperature sensitivity coefficient. The value range is limited to -0.05℃. -1 to -0.01℃ -1 The negative coefficient is determined based on the physical negative correlation characteristic that the viscosity of polymer base oil materials decreases monotonically with increasing ambient temperature.

[0073] Based on the aforementioned safety boundary construction for algorithm completeness, in order to prevent the calculated ambient temperature feedforward compensation coefficient from being affected by extreme high temperatures in the workshop... The deformation reversal becomes negative or zero, triggering a system deadlock in the lower-level decision-making process due to the global coupling dynamic convergence threshold. The data processing unit calculates the original ambient temperature feedforward compensation coefficient. Subsequently, a dynamic limiting mechanism based on multi-dimensional compatibility was forcibly applied. The data processing unit set a compliance upper limit threshold for the feedforward compensation coefficient. The lower limit for compliance is 1.5. The value is 0.5, determined based on the maximum permissible physical rheological distortion limit of the damping grease throughout its entire lifespan. When the original coefficients output by the formula deviate from the safe range of [0.5, 1.5], the logic judgment unit automatically activates the edge saturation pinning drive, feeding forward the ambient temperature compensation coefficient. The rigid constraint is applied to the nearest boundary extremum.

[0074] The technical purpose and physical meaning of this control strategy is that when the actual ambient temperature in the workshop is lower than the standard reference ambient temperature, the algorithm adaptively converts it into a compensation gain greater than 1, thereby amplifying the initial basic static convergence threshold inside the main control calculation module and compensating for the pseudo viscosity increase of the damping grease caused by the low temperature environment. Conversely, when the actual ambient temperature is higher, an attenuation coefficient less than 1 is adaptively generated to tighten the judgment boundary proportionally. Furthermore, the amplitude limiting mechanism ensures that the dynamic convergence judgment threshold of the entire coupled domain is always positive, thereby safely and robustly blocking the nonlinear influence of environmental thermal disturbances on the underlying friction steady-state judgment at the source of the control domain.

[0075] Specific application examples:

[0076] This application example describes an adaptive break-in process for controlling the rotational damping torque of the rotating shaft of a certain type of medical dialysis machine's articulated arm. The internal friction pair gaps of the dialysis machine's articulated arm are filled with high-viscosity perfluoropolyether damping grease.

[0077] Hardware configuration:

[0078] The main control computing module uses an industrial control computer with a real-time kernel, and the bus communication cycle is set to 2ms. The dynamic torque sensor in the sensing module has a range of 0–5 N·m and a stiffness of 1.2 × 10⁻⁶. 4 N·m / rad, synchronous sampling rate Set to 2000Hz, the cutoff frequency of the low-pass filter. The frequency was set to 150Hz. The angular velocity acquisition unit used a 23-bit high-resolution photoelectric absolute encoder. The environmental sensing module used a Pt100 platinum resistance industrial temperature sensor with a sampling frequency of 2Hz and a moving average filter time window width set to 20s. The temperature sensor was positioned 0.8m away from the center of the test spindle and equipped with a non-metallic anti-convection buffer cover.

[0079] Core control algorithm parameter configuration:

[0080] The standard reference ambient temperature for system calibration At 25℃, the damping grease viscosity temperature sensitivity coefficient -0.025℃ -1 Basic static convergence threshold The calibration value is 0.04. Thermal attenuation coefficient. The calibration is 4.5 × 10 -5 J -1 Sliding data window Set to 1.0s. This is the sampling period for first-order discrete-difference operations. Consistent with the motion loop control cycle, it is 2ms. Target tolerance zone. The values ​​are set to [1.8 N·m, 2.2 N·m]. The upper limit threshold for intervention is set to 4. =0.16. The decision period for the dual logic gate state latch. Set to 5 seconds, maximum work limit Set to 25000J.

[0081] Pre-diagnostic execution process:

[0082] In the specific execution process of this embodiment, the pre-diagnostic process controls the servo motor to... A diagnostic angular velocity of 1.0° / s drives the spindle fixture to perform... The initial static friction torque peak value was captured by a minute probe motion of 2.0°. The value is 3.65 N·m. The main control computing module is based on this. The values ​​are used for multidimensional matrix addressing to match and obtain the specific weight parameters of the current workpiece: low-frequency trend weight coefficient. =1.2, high-frequency fluctuation weighting coefficient =2.5.

[0083] Accordingly, the data processing unit calculates the dynamic convergence index by combining the derivative and standard deviation of the mean characteristic. The core application formula is:

[0084] ;

[0085] Experimental verification and effect comparison:

[0086] To verify the technical advantages of the present invention in eliminating environmental temperature difference interference and offsetting the false convergence of damping grease thermal decay, different workshop environmental temperatures were simulated in an environmental climate test chamber. The present invention scheme (experimental group) was compared with the conventional fixed threshold control scheme (control group) that does not have environmental feedforward and work decay compensation.

[0087] Low-temperature workshop environment test (simulating winter workshop environment, ambient temperature) =12℃)

[0088] Please see the appendix Figure 5 This figure visually illustrates the dynamic evolution of the underlying algorithm parameters of the control system during the break-in process at a low temperature of 12℃. In the figure, the X-axis (break-in time) represents the physical duration of the dialysis machine's articulated arm performing the rotational break-in action, measured in seconds; the Y-axis (convergence index amplitude) is a dimensionless algorithm index space used to calibrate the system's set judgment criteria against the convergence state actually monitored by the sensors.

[0089] Control group performance:

[0090] Due to the low initial ambient temperature, the initial dynamic viscosity of the damping grease is extremely high. During the initial break-in period, even with intense microscopic cutting at the microscopic peaks, the torque standard deviation and mean change rate rapidly decrease due to the low temperature and high viscosity. (The remaining text appears to be incomplete and requires further context.) Figure 5 In the middle, the dotted line (the control group has a fixed judgment threshold) The rigid threshold represented by the traditional method is a straight line parallel to the X-axis with a value of 0.04, completely ignoring the physical reality that the initial viscosity of the damping grease is extremely high at low temperatures. Because this fixed threshold is too low, the actual dynamic performance is masked by the high viscosity, causing the system to mistakenly believe that the break-in period is complete at 32 seconds (corresponding to the cross mark x in the diagram and the vertical dotted line), thus prematurely stopping the machine. After the workpiece is removed and allowed to stand at a standard temperature, the damping torque experiences a severe upward drift because the friction pair surfaces have not actually undergone sufficient cutting and break-in. The final static torque reaches as high as 2.55 N·m, significantly exceeding the tolerance limit of 2.2 N·m, resulting in a defective product.

[0091] Experimental group performance: The temperature acquisition unit obtained smooth real-time ambient temperature. =12℃, based on which the main control calculation module calculates the ambient temperature feedforward compensation coefficient:

[0092] ;

[0093] This coefficient did not trigger the amplitude limit. At this time, according to the formula... The initial global coupling dynamic convergence threshold is adaptively amplified to:

[0094] ;

[0095] In the appendix Figure 5 In the middle, the solid line represents the dynamic judgment threshold of the experimental group. The dotted line represents the core dynamic control boundary of this invention. At the 0s starting point, since the system recognizes a low temperature environment of 12℃, the feedforward algorithm comes into play, adaptively raising (relaxing) the judgment threshold, successfully compensating for the pseudo-high viscosity. The dotted line in the figure represents the dynamic convergence index of the experimental group. This represents the actual torque fluctuations and convergence state monitored and calculated in real time by the sensor. As the break-in time increases and the accumulated frictional work increases... As the temperature rises, heat buildup causes the damping grease to thin, resulting in an exponential, smooth decline in the solid line representing the judgment threshold. This actively tightens the judgment criteria, forcing the workpiece to undergo a deeper physical break-in process. The actual dynamic convergence index (dashed line) only truly falls below the extremely stringent dynamic judgment threshold (solid line) after thermal attenuation compensation at the 85th second (corresponding to the circled mark 'o' in the diagram, representing the true moment of this invention). It remains there for 5 seconds, signaling the completion of the process and ensuring sufficient break-in. Finally, after the workpiece is freed from thermal effects across the entire temperature range, the damping torque stabilizes at 1.98 N·m, fully conforming to the target tolerance zone and completely eliminating low-temperature pseudo-convergence.

[0096] High-temperature workshop environment and local dry friction resistance test (simulating summer workshop environment, ambient temperature) Comparison of long-term stability (at 38℃)

[0097] Control group performance:

[0098] At a high temperature of 38℃, the damping grease is initially too thin. Due to the lack of a feedforward tightening mechanism, the system experiences excessive internal shearing during continuous unidirectional rotation, causing frictional work to be converted into accumulated heat. At 110s, due to centrifugal and squeezing effects, the damping grease on the core friction surface is completely squeezed out, resulting in a change in the dynamic torque standard deviation. The time derivative undergoes a sudden abrupt change, triggering a dry friction state that is about to become stuck. However, the control group cannot identify the abrupt change rate of this high-frequency roughness noise and continues to force unidirectional rotation. Finally, at 145s, the accumulated work exceeds the limit, and irreversible macroscopic micro-protrusion cutting damage occurs on the surface of the friction pair, resulting in the direct scrapping of the workpiece.

[0099] Performance of the experimental group:

[0100] The algorithm first calculates the feedforward compensation coefficient for the original ambient temperature as follows:

[0101] ;

[0102] The initial judgment boundary is actively tightened to 0.675 × 0.04 = 0.027. When the operation reaches 105 seconds and the friction interface is about to experience dry friction jamming due to lack of lubrication, the logic judgment unit accurately detects that the instantaneous change rate of the high-frequency roughness noise exceeds the trigger limit, immediately stops the unidirectional rotation, and issues a lubricant redistribution control command. The drive execution unit instantly switches to an asymmetric micro-stroke reciprocating commutation motion mode (positive angular acceleration 1200° / s²). 2 Reverse angular acceleration 1200° / s 2 (The ratio of forward to reverse angular displacement is 1.5). The hydrodynamic pumping effect successfully forced the damping grease at the tolerance edge to flow back and penetrate into the core friction surface. After the high-frequency dispersion returned to stability, the system resumed unidirectional continuous rotation and finally triggered a precision stop at 160s with extremely stringent deep physical stability indicators. The final workpiece damping torque was 1.85 N·m, and the first-pass yield improved from 72.5% in the control group to 99.4%.

[0103] In-depth verification of long-term stability (in conjunction with appendix) Figure 6 ):

[0104] Please see the appendix Figure 6 The figure shows the time evolution trajectory of the static resistance torque (corresponding to scenario 2) when the workpiece is placed in a standard environment for static cooling after the break-in process is completed and the machine is stopped, directly verifying the actual results of the process. In the figure, the X-axis (static and long-term test time) represents the placement and inspection time after the process is completed, in minutes; the Y-axis (static damping torque of the joint arm) represents the actual physical feel torque of the final product, in N·m.

[0105] The two horizontal dotted lines in the figure (the boundary of the target torque tolerance zone) represent the extreme range of product qualification, with an upper limit of 2.2 N·m and a lower limit of 1.8 N·m. The area between these two lines is the absolutely safe qualification zone.

[0106] Thick dashed line (standard control group regimen):

[0107] This represents the trajectory of defective products produced using a traditional control scheme. Immediately after shutdown (around 0 minutes), the workpiece is heated, the damping grease is thin, and the torque appears to be within the acceptable range of around 2.0 N·m. However, as time passes and the workpiece cools, the rough surface from insufficient cutting combines with the restored high-viscosity damping grease, causing a severe upward drift in torque, ultimately exceeding the tolerance limit of 2.2 N·m, rendering the workpiece unusable.

[0108] Solid line (experimental group scheme of this invention):

[0109] The trajectory represents the superior product produced by the solution of this invention. Since the interference of thermal effects on the judgment boundary has been forcibly eliminated through the algorithm during the processing, ensuring that the micro-friction surface is fully cut, the actual torque remains very stable at about 1.98 N·m after the workpiece stops and no matter how much internal heat is dissipated, it is locked firmly in the center of the qualified zone. This strongly proves that the control system has fundamentally solved the resistance drift problem and achieved true deep physical steady state.

Claims

1. An automated control system for stabilizing the torque of a dialysis machine articulated arm, characterized in that, It includes a main control computing module with an internally allocated data processing unit and a logic decision unit, and a sensing and detection module, an environmental perception module, and a servo drive module connected to the main control computing module via an industrial data bus; The sensing and detection module includes a torque acquisition unit that synchronously acquires real-time torque data and an angular velocity acquisition unit that acquires real-time angular velocity data and absolute angular displacement data; the environmental sensing module includes a temperature acquisition unit that acquires real-time ambient temperature data. The servo drive module includes an instruction parsing unit and a drive execution unit with a servo motor, a reduction mechanism and a spindle fixture. The main control calculation module receives the real-time torque data, real-time angular velocity data, absolute angular displacement data, and real-time ambient temperature data. It constructs an ambient temperature feedforward compensation coefficient through the data processing unit, generates motion control commands through the logic judgment unit, and sends them to the command parsing unit. The drive execution unit drives the main spindle fixture tooling, which in turn drives the dialysis machine articulated arm to perform unidirectional continuous rotation or small angular displacement reversal motion.

2. The automated control system for stabilizing the torque of a dialysis machine articulated arm according to claim 1, characterized in that, The data processing unit is used to set a sliding data window, perform time-domain statistics on the real-time torque data within the sliding data window, and calculate the mean dynamic torque and standard deviation of the dynamic torque within the current sliding data window.

3. The automated control system for stabilizing the torque of a dialysis machine articulated arm according to claim 2, characterized in that, The data processing unit is used to perform multi-dimensional matrix mapping addressing in a preset internal parameter mapping table based on the pre-extracted initial static friction torque peak value, and to match and obtain the low-frequency trend weight coefficient, high-frequency fluctuation weight coefficient and thermal attenuation coefficient. The data processing unit is used to calculate and obtain the dynamic convergence index by combining the low-frequency trend weight coefficient, the high-frequency fluctuation weight coefficient, the absolute value of the first derivative of the dynamic torque mean with respect to time, and the dynamic torque standard deviation.

4. The automated control system for stabilizing the torque of a dialysis machine articulated arm according to claim 3, characterized in that, The data processing unit is used to perform continuous time integration on the real-time torque data and the real-time angular velocity data to obtain the cumulative frictional work.

5. The automated control system for stabilizing the torque of a dialysis machine articulated arm according to claim 4, characterized in that, The temperature acquisition unit is used to smooth the real-time ambient temperature data and output a smoothed real-time ambient temperature, which is received by the main control computing module. The data processing unit is used to combine the pre-calibrated damping grease viscosity-temperature sensitivity coefficient, the smooth real-time ambient temperature, and the pre-calibrated standard reference ambient temperature to calculate and construct the ambient temperature feedforward compensation coefficient. The logic determination unit is used to extract the pre-stored basic static convergence threshold with a value range of 0.01 to 0.10, and to calculate the global coupling dynamic convergence determination threshold in real time using the thermal attenuation coefficient, the cumulative friction work, the ambient temperature feedforward compensation coefficient, and the basic static convergence threshold.

6. The automated control system for stabilizing the torque of a dialysis machine articulated arm according to claim 5, characterized in that, The data processing unit is used to set a compliance upper limit threshold of 1.5 and a compliance lower limit threshold of 0.

5. When the ambient temperature feedforward compensation coefficient deviates from the safe range formed by the compliance lower limit threshold and the compliance upper limit threshold, the logic determination unit is used to activate the edge saturation pinning drive to rigidly limit the ambient temperature feedforward compensation coefficient to the nearest boundary extreme value.

7. The automated control system for stabilizing the torque of a dialysis machine articulated arm according to claim 6, characterized in that, When the average dynamic torque does not enter the set target tolerance zone and the dynamic convergence index is greater than the intervention upper limit threshold set to 3 to 5 times the basic static convergence threshold, the logic judgment unit is used to generate a speed reduction protection command and send it to the servo drive module to reduce the angular velocity of the servo motor according to the set slope. When the mean dynamic torque is within the target tolerance zone and the time derivative of the standard deviation of the dynamic torque exceeds the preset trigger limit, the logic determination unit generates a lubricant redistribution control command, instructing the servo drive module to stop unidirectional continuous rotational motion and switch to asymmetric micro-stroke reciprocating commutation motion.

8. The automated control system for stabilizing the torque of a dialysis machine articulated arm according to claim 7, characterized in that, When the average dynamic torque is within the target tolerance zone and the dynamic convergence index is less than the global coupling dynamic convergence judgment threshold and the duration of this condition reaches the set judgment period, the logic judgment unit is used to issue a process completion signal, instructing the servo drive module to execute the set maximum braking torque stop action allowed by the system and lock the current workpiece assembly state.

9. The automated control system for stabilizing the torque of a dialysis machine articulated arm according to claim 8, characterized in that, The drive execution unit is used to respond to the diagnostic frequency sweep command issued by the main control computing module, control the servo motor to drive the spindle fixture tooling, perform a single micro angular displacement detection motion, and cooperate with the sensing and detection module to capture the peak value of the initial static friction torque; The drive execution unit is used to respond to the lubricant redistribution control command, switch the servo motor to an asymmetric micro-stroke reciprocating commutation motion mode, and call the internally loaded preset forward commutation angle acceleration and reverse commutation angle acceleration.

10. The automated control system for stabilizing the torque of a dialysis machine articulated arm according to claim 1, characterized in that, The torque acquisition unit is equipped with a dynamic torque sensor. The input end of the dynamic torque sensor is connected to the output shaft of the reduction mechanism via a high-rigidity diaphragm coupling, and the output end of the dynamic torque sensor is connected to the main shaft fixture. The sensing module is used to utilize the distributed clock synchronization mechanism of the industrial data bus to synchronize and package the torque data channel and angular velocity data channel with the same clock timestamp.

Images