Absolute value encoder of photomagnetic dual-detection module collaborative layout structure

Abstract

The invention discloses an absolute value encoder of an optomagnetic dual-detection module collaborative layout structure, and relates to the technical field of position detection, the encoder enables an optical coded disc and a magnetic ring to be coaxially installed on a mechanical base, and an optical detection subsystem and a magnetic detection subsystem are collaboratively arranged around a detected rotating shaft; collecting a magnetic initial position on a calibration platform by taking an optical absolute position as a reference, and storing a magnetic compensation mapping and a signal quality threshold value into a nonvolatile memory; in the operation period, the signal processing unit obtains an optical absolute position and a compensated magnetic position, the sampling state is divided into a normal state, an optical degradation state, a magnetic interference state and a two-way abnormal state according to the position difference value and the optical and magnetic signal quality, and the optical absolute position or the compensated magnetic position is selected to be output accordingly. And a frozen position or a safe position is output in a double-path abnormal state, and a health state field is attached, so that integration of position measurement, self-diagnosis and function safety is realized.

Description

Technical Field

[0001] This invention relates to the field of position detection technology, specifically to an absolute encoder with a cooperative layout structure of optical and magnetic dual detection modules. Background Technology

[0002] With the increasing application of collaborative robots, semiconductor manufacturing equipment, and autonomous driving chassis systems, the detection of angles and displacements of moving parts has become the foundation for achieving closed-loop control and safety interlocking. In engineering practice, components such as motor shafts, joint output shafts, and slides often rely on absolute encoders to obtain position feedback, and existing solutions mainly use optical encoders or magnetic encoders. Optical encoders are widely used in high-precision applications due to their high resolution and small nonlinear error, but the optical path and code disk need to be in a clean and stable environment, and they are relatively sensitive to oil, dust, condensation, and vibration. Magnetic encoders have a simpler structure and stronger environmental adaptability, and are more commonly used in oil mist, dust, and vibration conditions, but their absolute accuracy and resolution are limited, and the magnetic field of the magnet is also easily affected by the stator current of the motor and the surrounding rigid structure, causing zero-position offset and linearity changes.

[0003] In complex industrial environments, equipment often operates at high speeds and accelerations for extended periods, surrounded by interference from oil mist, water vapor, dust, and electromagnetic fields. Optical encoders may experience signal attenuation, positional shifts, or even code loss due to contamination, condensation, or micro-vibrations on the transparent window or etched surface causing changes in the light spot shape. Magnetic encoders, on the other hand, may suffer from systemic errors that fluctuate with operating conditions due to changes in the magnetic circuit characteristics caused by variations in the external magnetic field. It is difficult to simultaneously achieve both accuracy and reliability.

[0004] In engineering practice, there have been attempts to stack optical encoders and magnetic encoders on the same drive shaft, or to install the two types of encoders relatively independently within the same housing. However, these combinations are mostly aimed at increasing the number of channels, and the structural layout usually does not take into account the synergistic characteristics of the two measurement principles. This can easily lead to differences in axial position and radial eccentricity, resulting in inconsistent measurement references. Simple stacking or side-by-side installation may also cause the circuitry and support structure of the optical module to change the surrounding magnetic circuit distribution. At the same time, the permanent magnet and its supporting structure encroach on the installation space and heat dissipation path of the optical module. Once the encoder used for position feedback experiences output instability or long-term drift under critical operating conditions, it will cause servo closed-loop error accumulation or even loss of control, resulting in workpiece scrapping, unplanned production line downtime, and safety accidents. Existing absolute encoder structures are difficult to simultaneously take into account optical and magnetic measurement principles and control the interaction between multiple physical fields within a limited installation space.

[0005] Therefore, the current technical problem is that in motion control systems with high requirements for accuracy, reliability and functional safety, existing absolute encoders are difficult to achieve a reasonable layout and stable coordination of optical and magnetic measurements within a limited installation space, and are difficult to simultaneously take into account the accuracy and environmental adaptability of the two types of sensing principles, thus making it difficult to provide a reliable position feedback basis for the upper control system. Summary of the Invention

[0006] (a) Technical problems to be solved To address the shortcomings of existing technologies, this invention provides an absolute encoder with a collaborative layout structure of optical and magnetic dual-detection modules. It acquires the initial magnetic position on a calibration platform using the optical absolute position as a reference, and stores the magnetic compensation mapping and signal quality threshold in a non-volatile memory. During operation, the signal processing unit acquires the optical absolute position and the compensated magnetic position. Based on the position difference and the quality of the optical and magnetic signals, the sampling state is divided into normal, optical degradation, magnetic interference, and dual-path abnormal states. Accordingly, it selects either the optical absolute position or the compensated magnetic position for output. In the dual-path abnormal state, it outputs a frozen position or a safe position along with a health status field, thus solving the technical problems described in the background art.

[0007] (II) Technical Solution To achieve the above objectives, the present invention provides the following technical solution: An absolute encoder with a dual optical and magnetic detection module cooperative layout includes a mechanical base, a test shaft rotatably mounted on the mechanical base, an optical code disk and a magnetic ring coaxially fixed to the test shaft, an optical detection subsystem disposed on the mechanical base and aligned with the optical code disk, a magnetic detection subsystem disposed on the mechanical base and aligned with the magnetic ring, a signal processing unit electrically connected to the optical detection subsystem and the magnetic detection subsystem, a non-volatile memory electrically connected to the signal processing unit, and a communication interface electrically connected to the signal processing unit. The non-volatile memory stores the mapping parameters for converting the initial magnetic position to the magnetic compensation position, as well as the signal quality threshold. The signal processing unit synchronously acquires the optical absolute position and the magnetic compensation position according to the sampling period during operation, calculates the difference between the two, and classifies the health status of the current sampling period into one of the following: normal state, optical degradation state, magnetic interference state, and dual-path abnormal state, based on the health status of the current sampling period. It selects or weights the optical absolute position and the magnetic compensation position to generate the position output according to the health status of the current sampling period. In the dual-path abnormal state, it outputs the frozen position or the safe position and outputs the health status information through the communication interface.

[0008] Furthermore, the optical detection subsystem and the magnetic detection subsystem are each an independent submodule and are fixed together on the same mechanical base. The mechanical base is provided with positioning surfaces, adjusting screws and positioning pins for adjusting the axial position and circumferential alignment of the two submodules relative to the measured shaft, so as to limit the assembly gap and keep the optical code disk and the magnetic ring coaxial.

[0009] Furthermore, the magnetic detection subsystem is arranged on the side of the optical detection subsystem facing away from the optical path, so that the optical detection subsystem and the magnetic detection subsystem form a vertical stacked structure in the axial direction of the shaft being measured. A conductive metal shielding plate electrically connected to the mechanical base is set between the two, and a grounding terminal is set on the mechanical base, so that the shielding plate and the external grounding conductor form a continuous shielding loop.

[0010] Furthermore, the non-volatile memory stores calibration data pairs corresponding to multiple reference position points. Each calibration data pair includes at least the optical position and the initial magnetic position corresponding to the same reference position point. The multiple reference position points cover the effective measurement range of the absolute encoder.

[0011] Furthermore, the non-volatile memory also stores several magnetic position compensation lookup tables according to angular intervals, and stores optical thresholds and magnetic thresholds determined by the range of values ​​of optical signal intensity, optical contrast and magnetic field amplitude.

[0012] Furthermore, the signal processing unit is configured to: calculate the difference between the optical absolute position and the magnetic compensation position in each sampling period, and statistically analyze multiple consecutive differences within a preset time window; at the same time, compare the optical signal intensity and magnetic field amplitude in the corresponding sampling period with the optical threshold and magnetic threshold, respectively, and thereby classify the health status of the current sampling period into one of the following: normal state, optical degradation state, magnetic interference state, and dual-path abnormal state.

[0013] Furthermore, the signal processing unit is configured to: generate position output in a manner where the optical absolute position weight is greater than the magnetic compensation position weight when the health status of the current sampling period is normal; gradually decrease the optical absolute position weight and gradually increase the magnetic compensation position weight according to a preset time sequence when the health status of the current sampling period is optical degradation; and generate position output using only the optical absolute position when the health status of the current sampling period is magnetic interference.

[0014] Furthermore, the signal processing unit is configured to: when the health status of the current sampling period switches from a non-dual-path abnormal state to a dual-path abnormal state for the first time, record the last position output before the switch as the frozen position, and repeatedly output the frozen position during the period of maintaining the dual-path abnormal state; when no valid frozen position is obtained, output a fixed safe position according to the pre-stored safe position parameters, and output health status information representing the dual-path abnormal state through the communication interface.

[0015] Furthermore, the signal processing unit is configured to: set a health status field in the data output through the communication interface, periodically update the current health status in the health status field, and simultaneously count the cumulative number of sampling cycles of optical degradation status and magnetic interference status, and output the current health status and cumulative occurrence count through the communication interface for the host system to make maintenance decisions.

[0016] Furthermore, the optical detection chip in the optical detection subsystem and the magnetic detection chip in the magnetic detection subsystem are mounted on the same circuit board. The circuit board is equipped with a dedicated integrated circuit chip, which integrates the optical detection circuit and the magnetic detection circuit. The relative positions of the optical sensitive area and the magnetic sensitive area are defined by the circuit board wiring and the layout design of the dedicated integrated circuit chip. The inherent alignment of the optical code disk and the magnetic ring is achieved by the cooperation of the package shell and the mechanical base.

[0017] (III) Beneficial Effects This invention provides an absolute encoder with a cooperative layout structure of optical and magnetic dual detection modules, which has the following advantages: By coaxially fixing the optical code disk or grating element with the magnetic ring or permanent magnet assembly on the mechanical base, and by coordinating the optical and magnetic subsystems around the same measured axis and isolating the optical path and magnetic flux by the shielding structure, a unified optical measurement surface and magnetic measurement surface are formed at the same angular position, providing a stable and consistent physical basis for optical-magnetic joint calibration and subsequent fusion processing.

[0018] By synchronously acquiring optical absolute position and magnetic initial position on a calibration platform according to a predetermined trajectory, a calibration relationship between the magnetic initial position and the compensated magnetic position is established based on the optical absolute position. This calibration relationship, along with the optical signal intensity and magnetic field amplitude threshold, is then stored in a non-volatile memory. This enables the magnetic subsystem to inherit the optical scale during operation and to independently undertake absolute position measurement when the optical subsystem degrades.

[0019] By synchronously acquiring the optical absolute position and the compensated magnetic position at a uniform sampling period during operation, and combining the comparison results of the optical signal intensity and magnetic field amplitude with the threshold, a health status mark is generated for each sampling period, and at least the normal state, optical degradation state, magnetic interference state and dual-path abnormal state are distinguished. This enables the encoder to continuously obtain diagnostic information that reflects the state changes of the optical subsystem and magnetic subsystem throughout its entire life cycle.

[0020] The signal processing unit selects the fusion method of optical absolute position and compensated magnetic position based on the health status marker. In the normal state, it outputs the optical-magnetic fusion absolute position. In the optical degradation state, it increases the weight of the compensated magnetic position. In the magnetic interference state, it shields the compensated magnetic position. In the dual-path abnormal state, it outputs the safe position and synchronously outputs operation and maintenance auxiliary information, so that the position output link and the health diagnosis link form a clear control relationship within the same encoder. Attached Figure Description

[0021] Figure 1 This is a schematic diagram of the absolute encoder structure of the optical-magnetic dual-detection module cooperative layout in an embodiment of the present invention.

[0022] Figure 2 This is a schematic diagram illustrating the working principle of the absolute encoder with a cooperative layout structure of optical and magnetic dual detection modules in an embodiment of the present invention. Detailed Implementation

[0023] The technical solutions of 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.

[0024] Please see Figure 1 and Figure 2 This invention provides an absolute encoder with a cooperative layout structure of optical and magnetic dual detection modules.

[0025] This absolute encoder includes a mechanical base, the measured shaft, an optical code disk, a magnetic ring, an optical detection subsystem, a magnetic detection subsystem, a signal processing unit, a non-volatile memory, and a communication interface. This absolute encoder achieves position measurement, self-diagnosis, and safe output through a cooperative layout structure, calibration and parameter fixing mechanisms, operational health status determination mechanisms, and position output and anomaly degradation control mechanisms. Its structural composition and working mechanism will be explained below.

[0026] I. Description of Collaborative Layout Structure Around the shaft being measured, a coordinated layout structure is formed on the mechanical base to stably support the optical detection subsystem and the magnetic detection subsystem, so that the two types of detection modules share the same angular reference geometrically. The interference between the optical path and the magnetic path is reduced through reasonable spatial layering and shielding design, while reserving reasonable path and installation space for the signal front-end circuit.

[0027] In typical scenarios such as collaborative robot joints and semiconductor device drive shafts, the measured shaft usually serves both torque output and position detection functions. If the optical detection subsystem and the magnetic detection subsystem are arranged based on different mechanical reference surfaces or different axes, even if the accuracy is acceptable when calibrated individually, the interpretation of the same actual rotation angle by the two measurement chains will slowly drift with temperature changes, stress release of the supporting structure, or long-term wear, causing subsequent optical-magnetic joint calibration to lose its reference value.

[0028] Therefore, it is necessary to precisely coaxially mount the optical code disk or grating element with the magnetic ring or permanent magnet assembly around a unified mechanical reference, and provide a reliable mounting surface for the optical and magnetic read heads on the same mechanical base, so that the two measurement chains are structurally coaxial and share a common reference.

[0029] The mechanical base serves as the platform supporting all internal components of the encoder. Its reference holes, positioning surfaces, and mounting surfaces must not only meet the rotational accuracy requirements of the measured shaft, but also provide a repeatable positioning reference for the optical and magnetic reading heads.

[0030] The tested shaft is supported on a mechanical base by paired bearings. The outer ring of the bearing forms the first positioning relationship with the mechanical base, and the inner ring of the bearing forms the second positioning relationship with the tested shaft. The optical code disk or grating element and the magnetic ring or permanent magnet assembly form the third positioning relationship with the tested shaft through connecting parts such as sleeves and pressure rings. The mounting surfaces of the optical and magnetic read heads further establish a fourth positioning relationship with the reference surface of the mechanical base, thus forming a multi-level geometric chain from the mechanical base to the tested shaft, and then to the optical and magnetic sensing elements.

[0031] In one embodiment, the mechanical base is integrally machined, with a through-hole machined at its center as the main support hole for the shaft under test. Two bearings are pressed into the two ends of the through-hole to ensure the roundness and coaxiality between the outer rings of the bearings and the mechanical base. The shaft under test passes through the inner rings of the two bearings in sequence and is fixed by a shoulder and a locking nut, thereby locking the shaft under test and the mechanical base in an axial position and forming a repeatable rotational motion with the rolling contact surface of the bearings as the fulcrum.

[0032] The optical code disk is fastened to the shoulder of the shaft under test, and the magnetic ring or permanent magnet assembly is fastened to another shoulder or outer circle of the sleeve near the optical code disk, so that the optical engraving circle and the magnetic pole distribution circle are both referenced to the shaft under test in terms of radius and axial position.

[0033] Geometrically, the radial deviation function can be expressed as: Among them, the radial deviation function : In the circumferential angle The radial difference between the lower optical etched circle and the magnetic pole distribution circle; used to describe the radial difference between the optical etched circle and the magnetic pole distribution circle at different azimuth angles, mainly used to explain mechanical design objectives and assembly verification methods. Since this function is indirectly implemented through machining tolerance control and sampling inspection, it does not require specific calculation in the firmware. optical engraving circle radius : In the circumferential angle The actual radius of the optical encoder disc engraving circle relative to the theoretical center of the measured shaft; the radius of the magnetic pole distribution circle. : In the circumferential angle The actual radius of the trajectory of the magnetic pole center of the magnetic ring relative to the same center of the circle. Circumferential angle. The range of values ​​is arrive , is used to describe the full angular position within a circumference.

[0034] Through machining and assembly control, the radial deviation function is made By maintaining the measured shaft within a predetermined allowable range across the entire angular range, the geometric responses of the optical and magnetic detection subsystems to changes in the measured shaft angle are ensured to be as consistent as possible. This allows on-site assembly personnel to sequentially complete actions such as mechanical base clamping, bearing press-fitting, insertion and locking of the measured shaft, and positioning and securing of the optical encoder and magnetic ring. After completion, mechanical gauges are used to randomly check the relative deviations at multiple angular positions on the optical engraving circle and magnetic pole distribution circle to confirm that they meet the requirements. By ensuring that both the optically and magnetically sensitive areas use the measured shaft as the sole reference, achieving structural coaxiality and a common reference, it helps reduce systematic errors that need to be compensated for during subsequent joint calibration, improving the overall feasibility of the solution.

[0035] In another implementation, using a collaborative robot joint as an example, the mechanical base is designed as a ring-shaped shell, with a hollow internal area for power cables and cooling pipes. For a discrete collaborative layout, two independent mounting platforms are provided on the outer circumference of the mechanical base: one for mounting an optical reader and the other for mounting a magnetic reader.

[0036] The optical reader is connected to the first mounting platform via an adjustment slot and fastening screws. On-site assemblers first align the center of the optical reader's field of view with the reference circle of the optical code disk's engraving, then tighten the fastening screws. The magnetic reader is connected to the second mounting platform via a pin positioning hole, with its sensitive surface facing the magnetic ring's pole distribution circle. Both mounting platforms share the same center and reference plane in their design, ensuring that after assembly, the optical and magnetic readers maintain a corresponding relationship with the measured shaft in both circumferential and axial positions.

[0037] In the integrated collaborative layout implementation, the optical detection chip and the magnetic detection chip are soldered together in the central area of ​​a high-density circuit board. Fixing holes are provided on the edge of the circuit board, and the entire circuit board is fixed to a single mounting surface within the mechanical base using screws. During circuit board layout, the optical detection chip is placed closer to the optical code disk, and the magnetic detection chip is placed closer to the magnetic ring. The two are connected by circuit board traces, eliminating the need for separate wiring harnesses. During assembly, only a single adjustment of the circuit board's planar position and angle relative to the mechanical base mounting surface is required to simultaneously determine the spatial positions of the optical and magnetic detection chips relative to the tested shaft, thus simplifying the on-site debugging process.

[0038] In use, the discrete collaborative layout provides sufficient mechanical adjustment freedom for each of the two detection modules, which is conducive to independent optimization of the optical and magnetic paths; the integrated collaborative layout pre-fixes the relative geometric relationship between the two types of sensitive areas during the chip and circuit board design stage, reduces variable factors in assembly, and makes the collaborative layout structure have good repeatability in mass production.

[0039] Furthermore, after the mechanical base and the measured shaft are coaxially arranged, the relative positions of the optical and magnetic detection subsystems need to be further defined within a limited space to minimize mutual interference while ensuring their respective operating conditions. The optical detection subsystem is sensitive to stray light, electromagnetic noise, and mechanical vibration, while the magnetic detection subsystem is more sensitive to the distribution of the external magnetic field and the arrangement of the magnetically conductive structure. If the optical and magnetic read heads are simply placed side by side or stacked without a comprehensive design of the shielding structure and signal path, problems may arise such as the influence of the light source driving circuit on the magnetic field, the attraction of the permanent magnet to the optical support components, and additional electromagnetic coupling introduced by the cable routing. Therefore, it is necessary to define the principle of coplanar arrangement or vertical stacking at the spatial organization level, and to design the shielding and signal front-end in conjunction with the specific configuration of the optical and magnetic detection subsystems.

[0040] In terms of spatial organization, coplanar arrangement facilitates observation of the geometric relationship between the optical and magnetic read heads and the measured shaft within the same axial section, which is beneficial for assembly and maintenance; vertical stacking can shorten the overall length in the direction of the measured shaft, making the encoder more suitable for installation in space-constrained joints or actuators.

[0041] Regardless of the method used, a shielding layer needs to be placed between the optically sensitive area and the magnetically sensitive area to attenuate high-frequency electromagnetic interference and low-frequency magnetic field coupling. The signal front-end circuit should be arranged near the shielding layer to make the lead between the sensitive element and the front-end amplifier circuit as short as possible.

[0042] In the coplanar arrangement embodiment, a ring-shaped lateral shielding plate is arranged circumferentially inside the mechanical base. This shielding plate is made of a highly permeable magnetic material, with its inner surface facing the magnetic ring and its outer surface facing the optical read head mounting area. During assembly, the magnetic read head is fixed on the inner platform of the lateral shielding plate, with its sensitive surface facing the magnetic pole distribution circle of the magnetic ring; the optical read head is fixed on the outer platform of the lateral shielding plate, with its field of view aligned with the optical code disk lines through a window opened on the mechanical base. Small holes are reserved on the lateral shielding plate for cables to pass through, and metal connections and grounding connections are made around the small holes to prevent the cables passing through the small holes from forming obvious magnetic and electromagnetic leakage channels. In the specific application scenario of collaborative robot joints, the assembler can remove the end cap along the joint shell to see that the optical read head, lateral shielding plate, and magnetic read head are arranged in an orderly manner around the measured shaft in the same plane. During maintenance, the detection module on one side can be disassembled and assembled without disrupting the geometry of the other side.

[0043] In the vertically stacked implementation, the optical detection subsystem is arranged on one side near the outer end of the shaft being tested, and the magnetic detection subsystem is arranged on the back of the optical detection subsystem, closer to the inner side of the mechanical base. A flat shielding layer is inserted between the two, which is fastened to the mechanical base with multi-point screws and grounded to form a continuous shielding surface.

[0044] The optical field of view observes the optical code disk through a transparent window in the axial direction, while the magnetic sensor senses the magnetic field of the magnetic ring through an axial or radial air gap. During on-site assembly, the operator first connects the shielding layer to the mechanical base, then installs the optical read head assembly on the outside of the shielding layer and the magnetic read head assembly on the inside of the shielding layer. The shielding layer acts as a natural separator, ensuring that the components on both sides do not interfere with each other during assembly.

[0045] Under this spatial organization and shielding structure, the attenuation relationship of the shielding layer to the magnetic field can be expressed by the magnetic field strength function after shielding: Among them, the magnetic field strength after shielding At a distance from the normal direction of the shielding layer Magnetic field strength at the location of the magnetic sensor; magnetic field strength before shielding. Magnetic field strength at the same location without a shielding layer; magnetic permeability of the shielding material. : The magnetic permeability characteristics of the high-permeability material used in the shielding layer; Shielding layer thickness This indicates the actual thickness of the shielding layer in the normal direction; , This is used to explain the attenuation mechanism of the magnetic field by the shielding layer. The expression is only in the typical exponential attenuation form. This kind of relationship is common in electromagnetic compatibility design. It is only used to illustrate the principle that increasing the permeability or thickness can enhance the shielding effect, and it does not require precise calculation according to the formula.

[0046] Exponential function It is a natural exponential function used to describe the exponential decay trend of electromagnetic fields in a homogeneous medium.

[0047] By selecting the magnetic permeability of the shielding material and shielding layer thickness It is possible to increase the magnetic field strength after shielding while maintaining a compact structure. The design closely approximates the expected operating range of the magnetic sensor, thus balancing magnetic measurement accuracy with isolation of the optical subsystem. The coplanar arrangement implementation facilitates on-site observation and maintenance, while the vertical stacking implementation reduces axial length and improves internal space utilization. In both implementations, the shielding layer separates the optically sensitive area from the magnetically sensitive area, keeping multi-physics coupling within a predictable range and providing a relatively stable electromagnetic environment for subsequent signal front-end design.

[0048] In the construction of the optical detection subsystem, the designers installed an optical code disk with multiple engravings in the middle of the shaft being tested, and the optical read head contains a light-emitting element, an imaging lens, and a photoelectric detection array.

[0049] As the optical code disk rotates with the shaft under test, the etched lines block and transmit the light beam, and each pixel on the photodetector array outputs a photocurrent related to its angular position. The signal front-end circuit is compactly arranged on a small circuit board behind the optical read head housing. The photodetector array is connected to the amplification and shaping circuit through short leads to reduce parasitic capacitance of the leads and external electromagnetic coupling. During assembly, the worker first fixes the optical code disk to the shoulder of the shaft under test, then adjusts the optical read head to a position with clear imaging through the adjustment slot, and tightens the screws to complete the assembly of the optical detection subsystem.

[0050] In the construction of the magnetic detection subsystem, the magnetic ring or permanent magnet assembly is fixed close to the optical code disk to maintain a compact overall length. The magnetic read head internally houses a differentially structured silicon magnetic transistor and a Hall element. The silicon magnetic transistor detects the magnetic field distribution of the magnetic ring in the radial and tangential directions, while the Hall element detects the magnetic field distribution in the axial direction. The signal front-end circuitry of the magnetic read head is located on a circuit board near the sensing element. Differential amplification and temperature compensation circuitry convert the weak magnetoelectric signal into a voltage signal with amplitude and linearity suitable for subsequent digital processing.

[0051] In one implementation, the signal front-end circuits of the magnetic and optical read heads are respectively arranged on both sides of the shielding layer, so that the optical front-end circuit mainly processes photocurrent, and the magnetic front-end circuit mainly processes magnetoelectric signals, with the coupling path between the two separated by the shielding layer. Electromagnetic field finite element analysis tools can be used to model the combined structure of the magnetic ring, mechanical base, and shielding layer to evaluate the influence of different shielding materials and thicknesses on the magnetic field distribution and determine the optimal installation location range for the magnetic sensor. Simultaneously, optical design software can be used to model the combination of the optical encoder, imaging lens, and photoelectric detection array to evaluate the impact of different fields of view and light source arrangements on the optical signal quality.

[0052] By feeding the above modeling results back into the mechanical structure and circuit board layout design, the spatial relationship between the optical detection subsystem and the magnetic detection subsystem within the same encoder can be quantitatively verified, shortening the prototype debugging cycle.

[0053] Thus, the optical detection subsystem achieves a short-path, high signal-to-noise ratio photoelectric conversion chain, while the magnetic detection subsystem achieves a compact and relatively independent magnetic field sensing chain. Under the constraints of the shielding layer and the mechanical base, the two form a dual-channel measurement structure that does not interfere with each other but is interconnected.

[0054] II. Calibration and Parameter Consolidation Process Description: The optical detection subsystem and the magnetic detection subsystem have achieved a coaxial cooperative layout as described above. However, the outputs of the two measurement chains are still independent of each other. If the correspondence between the optical reference position and the initial magnetic position is not established through a unified calibration condition, it is impossible to accurately determine the source of magnetic measurement deviation during operation, and it is also difficult to reliably compensate based on magnetic measurement when optical degradation occurs.

[0055] To this end, a reference trajectory covering the working range is first designed on the calibration platform. Then, optical and magnetic initial positions are simultaneously collected at each reference position, and these data are organized into an ordered data chain. Subsequently, based on the variation law of optical-magnetic difference with position, an optical-magnetic mapping function is constructed. Combined with the amplitude, contrast and stability of optical and magnetic signals under calibration conditions, a signal quality benchmark threshold for distinguishing between normal and abnormal states is extracted. Finally, the mapping function and threshold are solidified in the encoder's internal memory for use during operation.

[0056] The aforementioned collaborative layout structure satisfies the axis deviation index. Under the required conditions, the shaft under test is smoothly rotated from the starting angle to the ending angle through a representative calibration motion, so that the optical detection subsystem and the magnetic detection subsystem go through the entire stroke range and acquire paired data at several representative reference positions.

[0057] Only when the calibration trajectory covers the working range that may be involved in subsequent actual use, and the distribution of reference position points takes into account both uniformity and concentration at key positions, can we ensure that the optical-magnetic mapping function does not undergo large-scale extrapolation during actual operation, thus guaranteeing the reliability of the compensation results.

[0058] Therefore, the mechanical base and the shaft under test are first installed as a whole on a dedicated calibration platform. This calibration platform can be a turntable or a linear motion mechanism with high repeatability. Its drive shaft is rigidly connected to the shaft under test through a coupling or flexible connector, and an angle reference ruler or high-precision angle sensor is set externally to control and record the reference trajectory.

[0059] The signal processing unit is connected to the optical detection subsystem and the magnetic detection subsystem via connecting cables, and at the same time establishes a trigger relationship with the control system of the calibration platform, so that whenever the calibration platform moves to the preset reference position, the signal processing unit can simultaneously acquire the output signals of the optical detection subsystem and the magnetic detection subsystem.

[0060] First, the number of calibration points is set on the calibration platform. It generates a set of reference position sequences, which can be distributed as multiple equally spaced points throughout the working range, and several dense points are added near the mechanical limit.

[0061] Afterwards, the operator starts the calibration platform, causing the shaft under test to stop at each reference position in sequence. At each reference position, the calibration platform control system sends a sampling trigger signal to the signal processing unit. Upon receiving the trigger signal, the signal processing unit reads the digital decoding result of the optical detection subsystem as the optical reference position sequence. Simultaneously, the solution results of the magnetic detection subsystem are read as the initial magnetic position sequence. Among them, the calibration point number from arrive Through this strict one-to-one data collection method, a system based on calibration point numbers was established. A pair of data chains indexed by an index.

[0062] As an implementation method for collaborative robot joints, the production line is equipped with a separate calibration station. Workers fix the assembled encoders onto the calibration platform using a special fixture, connect the signal lines of the optical detection subsystem and the magnetic detection subsystem, and select a predefined full-circle calibration program on the control panel.

[0063] The calibration platform slowly rotates the shaft under test according to the program, pausing briefly at each reference angle position. Indicator lights on the control panel indicate that sampling is complete, and the optical reference position sequence is displayed. With magnetic initial position sequence The data is written into the temporary storage area inside the encoder. Throughout the process, the worker only needs to observe the indicator light status and the platform rotation to complete one calibration. After calibration, the encoder has a set of optical and magnetic position correspondence data that spans the entire circle.

[0064] Therefore, each reference position corresponds to a unique optical reference position and a unique magnetic initial position, and the number of calibration points... Optical reference position sequence and magnetic initial position sequence The original data foundation for constructing the optical-magnetic mapping allows subsequent deviation analysis and mapping function fitting to no longer rely on external records, but are entirely handled by the encoder's internal data structure.

[0065] After completing the calibration condition design and synchronous data acquisition, a sequence of optical-magnetic difference values ​​that can be used for subsequent analysis needs to be directly organized within the signal processing unit. This creates an intermediate variable that centrally characterizes the relationship between the optical reference position and the initial magnetic position during the calibration phase. Otherwise, each subsequent analysis would require reading the optical reference position sequence separately. and magnetic initial position sequence Further subtraction not only increases the number of calculation steps but also makes the data structure more fragmented, which is not conducive to storage compression and fast lookup within the encoder.

[0066] Therefore, after the signal processing unit completes the acquisition of the optical reference position and the initial magnetic position at a certain calibration point, it can immediately calculate their difference in the local computing unit and use this difference as the optical-magnetic difference sequence. One element is stored. Simultaneously, the original optical reference position sequence... and magnetic initial position sequence It will also be retained for subsequent fitting of the optical-magnetic mapping function and generation of the compensated magnetic position.

[0067] Therefore, in the calibration point number After the corresponding sampling is completed, the signal processing unit calculates the optical-magnetic difference sequence according to the following relationship: Where: Optical-magnetic difference sequence : Indicates the calibration point number At the corresponding reference position, the difference between the optical reference position and the initial magnetic position can be greater than 0 or less than 0; optical reference position sequence : Indicates the calibration point number At this point, the absolute position angle value calculated by the optical detection subsystem, based on optical measurement, covers one or more complete measurement circles of the encoder. Magnetic initial position sequence : Indicates the calibration point number At this location, the magnetic detection subsystem calculates the uncompensated absolute position angle value based on the three-dimensional magnetic field. The range of this value corresponds to the optical reference position sequence. They are identical or correspond to each other through a simple linear transformation.

[0068] Therefore, after receiving the calibration trigger signal, the signal processing unit first reads the decoding result of the optical detection subsystem and writes it into the optical reference position sequence. Then, the solution results from the magnetic detection subsystem are read and written into the magnetic initial position sequence. Then, a subtraction operation is performed again in the local processing unit to obtain the photomagnetic difference sequence. All three are stored in an internal buffer. This process is repeated once after each calibration point is completed, until all calibration point numbers are obtained. from Increase to .

[0069] As an example, at the calibration station of a collaborative robot joint, the operator can see the calibration point number on the control interface. Advancement, optical reference position sequence Magnetic initial position sequence And optical-magnetic difference sequence The continuously accumulating data, while not requiring specific numerical values, provides a clear indication that the calibration process is progressing as expected. After calibration, the three sequences are fully present within the encoder, preparing for subsequent bias analysis and mapping function fitting.

[0070] When used, the optical-magnetic difference sequence It is directly constructed during the calibration stage, which reduces redundant calculations in subsequent analysis and provides a clearer data perspective for the subsequent establishment of the optical-magnetic mapping function, so that the deviation distribution between the optical reference position and the magnetic initial position exists in a compact form within the encoder.

[0071] Regarding how to utilize optical reference position sequences Magnetic initial position sequence and optical-magnetic difference sequence Construct and expand a mapping function from the initial magnetic position to the compensated magnetic position.

[0072] The magnetic detection subsystem is affected by the shape of the magnetic circuit, temperature and process. Even if it is mechanically coaxial with the optical detection subsystem, there is still a repeatable nonlinear deviation between its output and the optical reference position. If this deviation is not solidified in the form of a function, it is difficult to accurately reconstruct the position value close to the optical reference during the operation phase when optical degradation or magnetic-assisted compensation occurs.

[0073] During processing, it is first assumed that the axis deviation index in the aforementioned cooperative layout structure has been controlled within the design allowable range. Therefore, the deviation between the optical reference position and the magnetic initial position mainly originates from the nonlinearity and scaling error of the magnetic measurement chain itself. The signal processing unit uses the magnetic initial position sequence... As the independent variable, the optical reference position sequence As the objective quantity, the photomagnetic mapping function is fitted. The mapping function is then used to generate a compensated magnetic position sequence at the calibration point. .

[0074] Therefore, a photomagnetic mapping function is introduced. In the calibration point number At the location, the compensated magnetic position sequence With magnetic initial position sequence The relationship between them is satisfied: Where: Compensated magnetic position sequence : indicates the use of optical-magnetic mapping function The magnetic position sequence obtained after compensating for the initial magnetic position should have values ​​as close as possible to the optical reference position sequence. ; Optical-magnetic mapping function : Represents a functional relationship with the initial magnetic position as input and the corresponding optical reference position as the target, used to reconstruct the angle close to the optical reference from magnetic measurements during operation; Magnetic initial position sequence : is the optical-magnetic mapping function The input independent variable.

[0075] In one optional implementation path, the optical-magnetic mapping function It can be approximated using a cubic polynomial: Where: Optomagnetic mapping function : Indicates when the input angle variable The output compensated angle value at a given initial magnetic position; angle variable. : Represents the initial magnetic angle as the independent variable, and its value range covers the sequence of initial magnetic positions. All possible values; polynomial coefficients 、 、 、 : Represents the optical-magnetic mapping function The polynomial parameters are fitted to ensure that the polynomial parameters are obtained at all calibration points. Optical reference position sequence Compensated magnetic position sequence The overall difference between them decreases.

[0076] The signal processing unit can solve for the polynomial coefficients using the least squares method combined with interpolation algorithms. , , , You can choose to implement the sparse matrix solver in an embedded environment or call a pre-built polynomial fitting program. Regarding the number of calibration points... In many cases, it is also possible to first determine the initial magnetic position sequence. The process involves segmenting the data and then fitting local polynomials within each segment to achieve a segmented optical-magnetic mapping function. .

[0077] In use, the initial magnetic position sequence is constructed through the above-mentioned optical-magnetic deviation analysis and mapping function. Converted into a compensated magnetic position sequence In the calibration point number At the corresponding positions, the optical reference position sequence Compensated magnetic position sequence The difference between them is significantly reduced, thus providing a basis for maintaining acceptable positional accuracy by relying on magnetic measurements during the operation phase when optical degradation occurs.

[0078] Furthermore, during the operation phase, it is necessary to distinguish whether the contamination is caused by the optical detection subsystem or the interference from the external magnetic field in the magnetic detection subsystem. Simply relying on the position difference is insufficient to determine the cause. Therefore, it is essential to identify the typical quality range of the optical and magnetic signals under normal operating conditions during the calibration phase so that these ranges can be used as a basis for judgment during the operation phase.

[0079] The signal processing unit acquires the optical reference position sequence. and magnetic initial position sequence Simultaneously, the light intensity, contrast, and level stability of the optical detection subsystem, as well as the output voltage amplitude, waveform stability, and temperature response of the magnetic detection subsystem, are recorded.

[0080] After calibration, bounded interval analysis was performed on these quality indicators to extract the upper and lower limits of each type of indicator under normal operating conditions, forming the optical signal quality reference threshold and magnetic signal quality reference threshold in this invention.

[0081] During calibration, the front-end circuitry of the optical detection subsystem provides measurements of light intensity and contrast. The signal processing unit then processes these values ​​around the calibration point number. Accumulated, forming a set of optical quality samples.

[0082] By calculating the maximum and minimum values ​​and leaving an appropriate safety margin, the optical signal quality reference threshold range is generated; the front-end circuit of the magnetic detection subsystem provides samples of magnetic field amplitude and output voltage stability, and the signal processing unit uses the same approach to generate the magnetic signal quality reference threshold range.

[0083] Subsequently, the signal processing unit will use the photomagnetic mapping function polynomial coefficients 、 、 、 The optical signal quality reference threshold and the magnetic signal quality reference threshold are written together into the encoder's internal non-volatile memory, and an index structure is established so that the corresponding parameters can be quickly read according to the current working mode during operation. After calibration, the encoder retains these parameters even after power is turned off and on again, and can be put into use without repeating the calibration.

[0084] In the specific application scenarios of collaborative robot joints, the production line can record the optical signal quality reference threshold and magnetic signal quality reference threshold of each joint batch in the production database during each joint's factory calibration for subsequent maintenance and traceability. When on-site maintenance personnel find that the operating data of a certain joint has been hovering at the edge of the reference threshold for a long time, they can arrange to clean the optical window or inspect the magnetic shielding structure without having to recalibrate the entire process.

[0085] III. Explanation of the Health Status Assessment Process During Operation After completing the aforementioned calibration and parameter solidification, the encoder enters the operation phase. During this continuous motion around the measured shaft, the outputs of the optical detection subsystem and the magnetic detection subsystem are simultaneously acquired to establish the optical reference angle, the magnetic compensation angle, and the corresponding difference and smoothing difference during operation. The health status of each sampling cycle is determined by combining the optical signal intensity and magnetic field amplitude.

[0086] Based on the established optical-magnetic compensation mapping function and signal quality threshold parameters, a sequence of operating period differences between the optical reference angle and the magnetic compensation angle, as well as a sequence of operating period parameters such as optical signal intensity and magnetic field amplitude, are established around the continuous motion process of the measured shaft. Through synchronous sampling and sliding analysis, the encoder state of each sampling cycle is classified and marked, enabling the encoder to identify normal state, optical degradation state, magnetic interference state, and dual-path abnormal state during operation, thereby providing state input for subsequent data fusion and functional safety control.

[0087] The optical reference angle sequence, magnetic compensation mapping function, optical signal intensity threshold, and magnetic field amplitude threshold reflect the ideal relationship under factory calibration conditions. However, in actual operation, the equipment faces various influencing factors such as temperature changes, mechanical vibration, oil deposits, and external magnetic field fluctuations. These factors can cause the outputs of the optical and magnetic detection subsystems to deviate from their calibrated states. If the encoder lacks continuous observation and analysis of the optical-magnetic difference and signal quality during operation, it will be impossible to determine whether a position deviation is a fluctuation within the normal error band or caused by optical contamination, magnetic field interference, or simultaneous abnormalities in both channels.

[0088] Therefore, the calibration results obtained during the aforementioned calibration and parameter solidification process must be incorporated into the sampling, analysis, and judgment processes during the operation phase, so that each output location data is accompanied by a corresponding health status judgment.

[0089] Specifically, a fixed sampling period is set through the signal processing circuit board to synchronously acquire the optical signal output by the optical detection subsystem and the magnetoelectric signal output by the magnetic detection subsystem. The optical reference angle function and the magnetic compensation angle function during operation are calculated, and the optical-magnetic difference function and its smoothed form during operation are constructed. Subsequently, these functions, as well as the optical signal intensity and magnetic field amplitude during operation, are used as inputs. By comparing the optical signal intensity threshold, the magnetic field amplitude threshold, and the allowable range of angle difference, each sampling moment is mapped as a discrete health status marker.

[0090] The motion of the measured shaft during operation is determined by the actual control task; it may be slow following or high-speed reciprocating. If the sampling rhythm is set arbitrarily, key changes will not be captured under certain operating conditions, while resources will be wasted under others. Therefore, it is necessary to establish an operating period angle function based on a fixed sampling period, combining the optical reference angle and the magnetic compensation angle, so that subsequent health status assessments have a unified timeline reference.

[0091] Therefore, the sampling period can be set in the signal processing circuit board. Using this cycle as a rhythm, sampling control signals are simultaneously sent to both the optical detection subsystem and the magnetic detection subsystem. The optical detection subsystem outputs the detector pixel position once per sampling cycle, and the signal processing circuit board converts this into an operational optical reference angle function based on a pre-calibrated scaling factor and the zero-position pixel position. The discrete values ​​are obtained; the magnetic detection subsystem outputs a three-dimensional magnetic field voltage signal once in each sampling period. The signal processing circuit board first calculates the original magnetic angle according to the pre-calibrated linearization relationship, and then obtains the operating magnetic compensation angle function through the magnetic compensation mapping function. The discrete values ​​of .

[0092] Furthermore, the number of samples within the signal processing circuit board is represented by discrete sampling time numbers, and the actual time is compared with the number. Through relationships Correlation. For each sampling number, the signal processing circuit board reads the current pixel position of the optical detection subsystem. Then, according to a predetermined linear relationship, it is converted into an optical reference angle for operation. Simultaneously, the raw magnetic angle currently output by the magnetic detection subsystem is read. And based on the polynomial mapping coefficients determined during the aforementioned calibration and parameter solidification process. 、 、 Calculate the magnetic compensation angle during operation .

[0093] As an example: In the operation scenario of a collaborative robot joint, the joint drive motor performs repetitive lifting and lowering actions according to the upper control command. The signal processing circuit board performs these actions in each sampling cycle. A single optical-magnetic sampling is triggered. As the joint moves, the optical reference angle... and magnetic compensation angle A pair of time-varying numbers are formed inside the circuit board. The operator can see two angle curves changing around the same trend on the diagnostic interface, representing a sequence of changes.

[0094] When in use, the optical reference angle function during operation and the magnetic compensation angle function during operation At a unified sampling rhythm The following structure was constructed, making it possible to compare the optical and magnetic dual channels on the time axis, providing a complete and continuous data foundation for the subsequent calculation of the optical and magnetic angle difference sequence.

[0095] Furthermore, the optical-magnetic angle difference value of a single sample during operation is affected by both hardware noise and instantaneous mechanical disturbances. If the health status is judged solely based on a single difference value, it is easy to misjudge short-term local disturbances as long-term offsets. Therefore, it is necessary to construct an optical-magnetic angle difference value sequence during operation and form a smooth difference value sequence through recursion, so that the health status judgment can be sensitive to long-term trends without overreacting to instantaneous fluctuations.

[0096] First, at each sampling number Calculate the optical-magnetic angle difference. Then, the difference sequence is smoothed using a first-order recursive filter to generate a smoothed difference sequence filter. .

[0097] The optical-magnetic angle difference can be defined as the difference between the optical reference angle and the magnetic compensation angle during operation: Among them, the difference between optical and magnetic angles : In the sampling number The difference between the optical reference angle and the magnetic compensation angle at the corresponding moment; the optical reference angle during operation. Indicates the sampling number The angle value is calculated by the optical detection subsystem; the magnetic compensation angle during operation. Indicates the sampling number The angle value is obtained by the magnetic detection subsystem through a compensation mapping function.

[0098] To mitigate the impact of instantaneous fluctuations, a smoothed difference sequence can be constructed using a recursive relationship: Among them, smoothed difference Indicates the sampling number The smoothed angle difference used for health status assessment; the smoothed difference value of the previous time step is filtered. Indicates the sampling number Smoothing difference at the location; Optical-magnetic angle difference The current sampling number The original difference; the smoothing coefficient A dimensionless parameter between 0 and 1, used to determine the weights of historical and current values ​​in the smoothed sequence. The closer the smooth sequence is to 1, the more it emphasizes historical trends. The closer a smooth sequence is to 0, the more it emphasizes the current change.

[0099] In one embodiment of a semiconductor wafer transport device, during the repeated extension, retraction, and rotation of the transport arm, short-term torsional impacts occur due to changes in the frictional force of the wafer chuck. The signal processing circuit board calculates the optical-magnetic angle difference in each sampling cycle. Then, following the above recursive relationship, the smoothed difference is obtained. On-site engineers can observe on the monitoring software interface that the original difference curve has many sharp peaks, while the smoothed difference curve is flatter and can better reflect the long-term offset trend.

[0100] When using it, construct the optical-magnetic angle difference sequence during operation. Filtering with smoothed difference sequences This allows subsequent health status assessments to no longer rely on a single sampling, but rather on the trend of difference changes over a period of time, thus enabling the ability to distinguish between long-term offsets and short-term disturbances during operation.

[0101] However, the optical-magnetic angle difference alone cannot determine whether the deviation originates from the optical or magnetic detection subsystem. For example, an increase in the optical-magnetic angle difference at a certain moment could be due to oil contamination of the optical encoder disk, weakening the optical signal intensity, or fluctuations in the external magnetic field causing magnetic measurement offset. To achieve accurate health status classification, the angle difference sequence needs to be combined with quality sequences such as optical signal intensity and magnetic field amplitude, and compared with the optical signal intensity threshold solidified during the aforementioned calibration and parameter solidification process. The quality of the optical-magnetic signal at the current sampling moment is determined by the magnetic field amplitude threshold.

[0102] Therefore, the signal processing circuit board has each sampling number In addition to calculating the optical reference angle, With magnetic compensation angle In addition, it also records the intensity of optical signals. and magnetic field amplitude .

[0103] Optical signal strength The magnetic field amplitude can be represented by the photocurrent amplitude or corresponding voltage output from the front-end circuit of the optical detection subsystem. It can be obtained by vector synthesis of three-dimensional magnetic field signals. Then, the signal processing circuit board will process the optical signal intensity. With optical signal intensity threshold In comparison, the magnetic field amplitude Compared with the magnetic field amplitude threshold, combined with the smoothed difference The size of the sample determines the quality assessment result at the current sampling moment.

[0104] Further, the quality assessment shall be conducted in the following order: First, determine the optical signal strength. Is it below the optical signal strength threshold? If in multiple consecutive sampling periods All below If this is the case, it can be determined that there is contamination in the optical channel or attenuation of the light source; then the magnetic field amplitude is determined. Does it significantly deviate from the typical range of the calibration phase, for example, approaching the upper or lower limit of the magnetic field amplitude threshold? If in multiple consecutive sampling periods... If the magnetic channel exhibits violent fluctuations, it can be concluded that the magnetic channel is disturbed by an external magnetic field or that the magnet is unstable.

[0105] As an example of a collaborative robot joint, after the joint has been running on a production line for an extended period, lubricating oil mist inside the joint may gradually adhere to the optical window, affecting the optical signal strength. It will slowly decrease and eventually fall below the optical signal strength threshold. And magnetic field amplitude It remains within the normal range determined during the calibration phase, at which point the signal processing circuit board can be identified as exhibiting an optical degradation trend. Meanwhile, in another embodiment, when a high-current welding device is added near the joint, the magnetic field amplitude during the welding operation... There are spikes in the optical signal intensity at certain time intervals. If the situation remains relatively stable, it can be identified as a magnetic interference trend.

[0106] Therefore, each sampling period is explicitly compared with its respective threshold, enabling the signal processing circuit board to distinguish between optical channel deviation and magnetic channel deviation, providing source information for subsequent health status classification, thus avoiding the need to rely solely on the angle difference itself to make a judgment.

[0107] Because the health status information during the operation phase needs to be provided to the subsequent data fusion strategy and the upper control system in a concise and clear form, rather than being output directly in the form of multi-dimensional continuous quantities.

[0108] Therefore, a health status determination mapping needs to be defined in the signal processing circuit board to filter out the smooth angle difference. Optical signal strength With magnetic field amplitude In summary, this is mapped to discrete health status labels. This ensures that each sampling period corresponds to a state value that can be used for logical judgment.

[0109] Therefore, the health status determination mapping is defined as a multi-input discrete mapping, which smooths the differences of input variables. Optical signal strength and magnetic field amplitude As the independent variable, the output health status label will be used. As the dependent variable, health status label. It can take four integer values, corresponding to the normal state, optical degradation state, magnetic interference state, and dual-path abnormal state, respectively.

[0110] The health status determination mapping can be formalized as a functional relationship: Among them, health status markers : In the sampling number Encoder operating status category; determination mapping This represents a decision function that maps smoothed differences to signal quality parameters as state labels; Status label definition: : Normal state; Optical degradation state; Magnetic interference status; : Dual-path abnormal state.

[0111] Smoothing Difference : In the sampling number The optical-magnetic angle difference used for determination; optical signal intensity : In the sampling number The intensity of the optical signal observed at the location; the amplitude of the magnetic field. : In the sampling number The combined amplitude of the magnetic field observed at the location.

[0112] Specifically, health status markers The judgment rule can be expanded as follows: when the smoothed difference Not greater than the upper limit of the allowable difference And optical signal strength Located in the normal optical range Internal and magnetic field amplitude Located in the normal magnetic field range Internal time, definition When in recent times At least one of the samples Each sample satisfies ,and Still located Internal time, definition ;when lie in Inside, and Peaks appear, persistent deviations or exceedances occur When, define ;when ,and and When both exceed their respective normal ranges, define .

[0113] in, This is the upper limit of the allowable difference set based on the calibration. and This represents the boundary of the normal range for optical signal intensity. and This represents the boundary of the normal range for magnetic field amplitude; a long-term passage through a length of The number of samplings within the sliding window that meet the conditions is determined, where the threshold number of times the conditions are met is... .

[0114] The above-mentioned determination mapping is executed according to the aforementioned conditions in the implementation to obtain the health status label corresponding to the current sampling number. .

[0115] As a preferred embodiment, the signal processing circuit board calculates a health status flag in each sampling cycle. Then, the flag is written to the output register along with the current output angle value. The encoder transmits the angle value and the health status flag to the upper control system via the bus. When the upper control system receives a status flag indicating optical degradation, it can schedule a shutdown to clean the optical window in advance; when it receives a status flag indicating magnetic interference, it can schedule a diagnostic program to check peripheral equipment; when it receives a status flag indicating dual-path abnormality, it can immediately execute an emergency shutdown or enter a safe operating mode.

[0116] When used, a continuous smoothed difference sequence is filtered. and signal quality sequence , Summarized as discrete health status marker sequences This enables the output of health status to be in a form that can be directly judged by logic and called by the higher-level system.

[0117] IV. Position Output and Abnormal Degradation Control Instructions Using the health status tags obtained during the aforementioned operational health status assessment process, the optical absolute position during the operational period is determined. Magnetic compensation position and comprehensive health assessment value An output decision chain is introduced to form an executable set of fusion and degradation rules within the encoder: under normal conditions, it fully utilizes optical precision and magnetic robustness; under optical degradation conditions, it relies on magnetic compensation to maintain acceptable accuracy; under magnetic interference conditions, it takes optical absolute position as the main output; and under dual-path abnormal conditions, it enters a safe mode and outputs operation and maintenance alarm information, thereby providing the upper system with complete feedback that includes both position and health information.

[0118] A sampling number has already been obtained. After assigning the corresponding health status labels (normal, optical degradation, magnetic interference, dual-path abnormality), if the encoder still only outputs the raw position value of a single path, it cannot flexibly utilize the optical absolute position under different states. With magnetic compensation position Each has its own advantages. Therefore, it is necessary to construct a fusion output mechanism oriented towards health status tags within the signal processing unit, so that the encoder can perform optical-magnetic coordination when in good health, automatically implement degraded operation when a single channel degrades, and quickly cut off dangerous outputs when unavailable.

[0119] The signal processing unit uses the health status tag obtained during the aforementioned operational health status determination process as a driving force to process each sampling number. The corresponding fusion weight coefficient is selected based on the label: when the label is normal, the optical weight is set to be larger and the magnetic weight to be smaller, so that the fusion position is output. It is essentially equivalent to optical absolute position, but retains the suppression effect of magnetic link on dynamic error; when the tag is optically degraded, the optical weight is gradually reduced and the magnetic weight is increased, so that the fused position output gradually moves towards the magnetically compensated position. Migration enables smooth degradation; when the tag is subject to magnetic interference, the optical weight is directly reset to 1 and the magnetic weight is reset to 0 to avoid disturbance from the magnetic link; when the tag is subject to dual-path anomaly, the output is no longer based on the fusion formula, but instead the safe mode output strategy is called.

[0120] To achieve continuous and controllable optical-magnetic fusion output under normal, degraded, and magnetically interfered conditions, the fusion angle during operation can be defined in the signal processing unit. for: Among them, the integration perspective : In the sampling number The final angle value output at the corresponding moment is used for closed-loop control or position monitoring; optical weights : In the sampling number Optical absolute position in time-fusion output The percentage ranges from 0 to 1. Magnetic weight : In the sampling number Magnetic compensation position in time-fusion output The proportion, ranging from 0 to 1; optical absolute position The optical detection subsystem is used for sampling numbering. The angle value obtained at each moment; Magnetic compensation position The compensation function formed by the initial angle of the magnetic detection subsystem during the aforementioned calibration and parameter solidification process. The angle value obtained by correction under the action; optical weighting under normal and single-path degradation states. With magnetic weight The sum is constrained to 1 to maintain the convex combination of the fusion result between the two measurement chains.

[0121] in: Comprehensive health assessment score : indicates the first The overall health status evaluation result at each sampling time is used to characterize the comprehensive deviation of the current encoder from the normal calibration state; the larger the value, the more obvious the deviation of the current position measurement link from the calibration state. Sampling sequence number : indicates the first step in the process. Each sampling point can also be understood as a time marker numbered sequentially according to the sampling period, used to distinguish health evaluation results at different times; Difference sliding index : indicates the first The sliding statistic of the difference between the optical position and the compensated magnetic position at each sampling time is used to reflect the consistency of the position results of the optical and magnetic paths over a period of time; the larger the value, the more obvious the positional deviation between the optical link and the magnetic link. Difference weighting coefficient : Indicates the moving average index of the difference The weight in the comprehensive health evaluation value is used to adjust the strength of the influence of the optical-magnetic position deviation on the final health evaluation result; this coefficient is a non-negative constant. Optical quality indicators : indicates the first The normalized result of the optical signal quality at each sampling time is usually used to characterize the degree of deviation of the current optical signal intensity, contrast, or overall quality from the normal calibration state; if it is defined as a normalization quantity, it usually takes a value close to 1 under normal calibration conditions. Optical weighting coefficient : This represents the contribution weight of optical quality indicators to the overall health evaluation value when they deviate from the normal state, and is used to adjust the proportion of optical link anomalies in the overall evaluation; this coefficient is a non-negative constant; Magnetic quality indicators : indicates the first The normalized result of the magnetic signal quality at each sampling time is usually used to characterize the degree of deviation of the current magnetic field amplitude, stability or overall quality from the normal calibration state; if it is defined as a normalization quantity, it usually takes a value close to 1 under normal calibration state. Magnetic weighting coefficient : This represents the contribution weight of the magnetic quality index to the overall health evaluation value when it deviates from the normal state, and is used to adjust the proportion of magnetic link abnormalities in the overall evaluation; this coefficient is a non-negative constant; Therefore, the health status labels obtained during the aforementioned operational health status determination process can be mapped to a preset weight combination within the signal processing unit: when the label is normal and the comprehensive health evaluation value is... At low levels, optical weights Set to close to 1, magnetic weight Set to near 0; when the tag begins to degrade optically but has not yet entered a dual-path anomaly, in multiple sampling numbers Gradually decrease the optical weight by pressing the fixed step size. Increase magnetic weight , to make the fusion angle The approach has gradually shifted from primarily relying on absolute optical position to primarily relying on magnetically compensated position; when the tag experiences magnetic interference, the optical weighting is directly applied. Set as 1, magnetic weight Set it to 0 to prevent the magnetic link from affecting the output.

[0122] As an example of a collaborative robot joint, when the joint operates in a clean environment for an extended period and its health status label remains normal, the signal processing unit processes signals at each sampling number. Optical weights The magnetic weight is fixed to a constant slightly less than 1. Fixed to a constant slightly greater than 0, so that the fusion angle Almost the absolute position of the optical system The fusion angle is achieved by overlapping the labels while retaining a small amount of magnetic information to suppress transient errors caused by high-speed vibrations. When on-site maintenance personnel observe the health status label gradually changing from normal to optical degradation, the fusion angle can be seen on the host computer interface. The change curve gradually approaches the magnetic compensation position. However, no obvious mutations were observed.

[0123] Through this health-state-driven fusion angle calculation, the technical effect is that, on the one hand, the output under normal conditions fully utilizes the absolute optical position. It leverages the high precision advantage and improves dynamic robustness with minimal magnetic compensation; on the other hand, it adjusts optical weights to mitigate optical degradation and magnetic interference. With magnetic weight The proportion makes the fusion angle It can smoothly complete the conversion from optical-dominated to magnetic-dominated or purely optical output, avoiding the control shock caused by the one-size-fits-all switching in traditional solutions.

[0124] Furthermore, during the long-term operation of the encoder, the degradation of both optical and magnetic links often exhibits a gradual process, and changes in health status tags may cause short-term jitter. If the weights are switched immediately upon tag change, the fusion angle... It will frequently switch between different modes, which is not conducive to the stability of the servo system.

[0125] Therefore, a gradual degradation and status maintenance mechanism needs to be introduced when health status labels change, linking the health status label sequence with the comprehensive health assessment value. This combination forms a weight adjustment process with lag characteristics.

[0126] The signal processing unit at each sampling number This section records the current health status label and the comprehensive health evaluation value within the last few sampling numbers. Changes. When a health status label is detected to change from normal to optical degradation, the optical weight is not immediately adjusted. Instead of decreasing to a predetermined degradation value, the degradation is gradually reduced proportionally within a pre-defined degradation range; only when the comprehensive health assessment value... A full adjustment of the weights is only completed when the health status label continues to rise and exceeds a certain duration threshold. Conversely, when the health status label recovers from optical degradation to normal, the optical weights are also increased slowly. In this way, the fusion angle Smooth regression to the optically dominant state.

[0127] In one embodiment of a semiconductor device rotating platform, after prolonged operation, the optical window gradually becomes covered by tiny particles, and the health status label changes from normal intermittent to optical degradation, but the overall health evaluation value... It only rises slightly for a short period of time. At this point, the signal processing unit maintains the fusion angle according to the degradation interval rule. Still based on optical absolute position The output is primarily controlled, without immediately switching to magnetic dominance; when maintenance personnel have not performed cleaning operations, the overall health evaluation value... The encoder automatically transitions from optically dominated output to magnetically dominated output only when the output continues to rise over multiple sliding time windows, and attaches a health status field in the bus message that represents optical degradation and output degradation.

[0128] This gradual degradation and state preservation mechanism reduces the impact of short-term jitter of health status labels on the fusion angle. To mitigate the impact of frequent switching and avoid servo shocks, and to ensure that only services with a comprehensive health evaluation value are switched on. A thorough degradation action is only taken when there is a sustained deterioration, thereby extending the availability of high-precision output as much as possible while ensuring functional safety.

[0129] When the health status label indicates dual-path abnormality, it means the optical absolute position is abnormal. With magnetic compensation position None of them are consistently reliable; continue to output angles. This could directly lead to malfunctions in the servo system. In addition, for ease of maintenance, it is also necessary to output the health status label, optical degradation count, magnetic interference duration, and other information to the host system on the bus.

[0130] Therefore, it is necessary to build a set of safe mode output logic and health information external release logic inside the encoder, so that the encoder can quickly cut off dangerous output when it detects an unrecoverable anomaly, and provide data basis for fault location and predictive maintenance.

[0131] Therefore, when the health status tag is determined to be a dual-path abnormality and meets a certain duration condition, the signal processing unit stops outputting the angle according to the fusion formula, and instead outputs the frozen angle or the preset safety angle. At the same time, during normal operation, the signal processing unit continuously accumulates the optical degradation event count, the magnetic interference event count, and the sampling number range for each abnormality, and adds the health status field and event count field to the bus message, so that the host system can infer the long-term changes in the encoder's working environment from this information.

[0132] In the case of dual-path abnormality, in order to avoid sudden changes in the output angle causing impact on the servo system, the signal processing unit usually uses a combination of angle freezing and safety limiting to process the output.

[0133] When multiple consecutive sampling numbers of health status tags are detected When there is a dual-path anomaly, the signal processing unit first records the last fusion angle that is still considered acceptable. This is used as the frozen angle, and the frozen angle is continuously output in subsequent sampling until the host system performs a shutdown or power-on reset. For some applications that require outputting a fixed safe position in case of failure, a certain angle value can be predefined as the safe angle during the design phase. When entering a dual-path abnormal state and meeting the conditions, the signal processing unit directly switches the output to the safe angle.

[0134] As an example for a collaborative robot workstation, when the encoder's optical window becomes severely contaminated due to prolonged lack of maintenance and the magnetic link is simultaneously subjected to interference from a strong external magnetic field, multiple consecutive sampling numbers of the health status tag are detected. All were marked as dual-path anomalies, at which point the signal processing unit stopped updating the fusion angle. Instead, it remains at the last credible angle. If the robot controller detects a dual-path abnormality in the health status field and that the angle remains unchanged for an extended period, it will immediately stop the current task and release the joint torque to prevent the robot from continuing to move in an unknown angle state.

[0135] By using the freezing angle and safety limiting logic, it can be ensured that no unpredictable erroneous angles are input to the servo system under dual-path abnormal conditions, reducing the risk of secondary accidents caused by encoder failure. At the same time, by outputting the freezing angle or preset safety angle, a clear fault response basis is provided to the upper system, enabling the entire control system to meet the functional safety standard requirements for detectable faults and deterministic responses.

[0136] Under both normal and abnormal conditions, angle values ​​alone are insufficient to fully reflect the encoder's operational status, making it impossible for the host system to conduct targeted maintenance planning. Therefore, a health status field and an operation and maintenance assistance field are added to the external communication messages, allowing location data and health information to be output together.

[0137] Among them, the signal processing unit in each sampling number The health status label is encoded as a health status field. The operation and maintenance information, such as the cumulative count of optical degradation events, the cumulative count of magnetic interference events, and the sampling number length of the most recent continuous abnormality, are packaged into the message and output in the form of an additional register or extended data segment without changing the basic structure of the bus protocol.

[0138] Similarly, in the implementation of collaborative robot joints, the angle values ​​and health status fields of the joint encoders can be viewed in real time via host computer monitoring software. When the health status field remains normal for an extended period while the optical degradation event count increases slowly, engineers can schedule cleaning of the joint interior during production breaks. When the magnetic interference event count suddenly increases at a certain time, it can be checked whether any new high-current devices have been added to the site or whether existing shielding structures have been removed, allowing for timely layout adjustments. For long-running semiconductor equipment, the maintenance team can also compare the usage environment differences between different batches of equipment based on the abnormal duration sequence recorded in the messages, extracting experience beneficial for improving on-site processes and shielding design.

[0139] By publishing health status fields and operation and maintenance information through the bus, the analysis results of the encoder's internal optical link, magnetic link, and comprehensive health evaluation are made explicit. This allows the upper system to make maintenance decisions and risk assessments without directly accessing the underlying measurement details. At the same time, these operation and maintenance fields provide a data foundation for subsequent improvements to the collaborative layout structure of the optical and magnetic dual detection modules (including shielding schemes, layout schemes, and compensation strategies).

[0140] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. An absolute encoder with a cooperative layout structure of optical and magnetic dual detection modules, characterized in that: include, A mechanical base, a test shaft rotatably mounted on the mechanical base, an optical code disk and a magnetic ring coaxially fixed to the test shaft, an optical detection subsystem mounted on the mechanical base and aligned with the optical code disk, a magnetic detection subsystem mounted on the mechanical base and aligned with the magnetic ring, a signal processing unit electrically connected to the optical detection subsystem and the magnetic detection subsystem, a non-volatile memory electrically connected to the signal processing unit, and a communication interface electrically connected to the signal processing unit. The non-volatile memory stores the mapping parameters for converting the initial magnetic position to the magnetic compensation position, as well as the signal quality threshold. The signal processing unit synchronously acquires the optical absolute position and the magnetic compensation position according to the sampling period during operation, calculates the difference between the two, and classifies the health status of the current sampling period into one of the following: normal state, optical degradation state, magnetic interference state, and dual-path abnormal state, based on the health status of the current sampling period. It selects or weights the optical absolute position and the magnetic compensation position to generate the position output according to the health status of the current sampling period. In the dual-path abnormal state, it outputs the frozen position or the safe position and outputs the health status information through the communication interface.

2. The absolute encoder according to claim 1, characterized in that, The optical detection subsystem and the magnetic detection subsystem are each an independent submodule and are fixed together on the same mechanical base. The mechanical base is provided with positioning surfaces, adjusting screws and positioning pins for adjusting the axial position and circumferential alignment of the two submodules relative to the measured shaft, so as to limit the assembly gap and keep the optical code disk and the magnetic ring coaxial.

3. The absolute encoder according to claim 2, characterized in that, The magnetic detection subsystem is arranged on the side of the optical detection subsystem facing away from the optical path, so that the optical detection subsystem and the magnetic detection subsystem form a vertical stacked structure in the axial direction of the shaft being measured. A conductive metal shielding plate electrically connected to the mechanical base is set between the two, and a grounding terminal is set on the mechanical base, so that the shielding plate and the external grounding conductor form a continuous shielding loop.

4. The absolute encoder according to claim 1, characterized in that, The non-volatile memory stores calibration data pairs corresponding to multiple reference position points. Each calibration data pair includes at least the optical position and magnetic initial position corresponding to the same reference position point. The multiple reference position points cover the effective measurement range of the absolute encoder.

5. The absolute encoder according to claim 4, characterized in that, The non-volatile memory also stores several magnetic position compensation lookup tables according to angle intervals, and stores optical thresholds and magnetic thresholds determined by the range of values ​​of optical signal intensity, optical contrast and magnetic field amplitude.

6. The absolute encoder according to claim 5, characterized in that, The signal processing unit is configured to: calculate the difference between the optical absolute position and the magnetic compensation position in each sampling period, and statistically analyze multiple consecutive differences within a preset time window. At the same time, it compares the optical signal intensity and magnetic field amplitude in the corresponding sampling period with the optical threshold and magnetic threshold, respectively, and classifies the health status of the current sampling period into one of the following: normal state, optical degradation state, magnetic interference state, and dual-path abnormal state.

7. The absolute encoder according to claim 6, characterized in that, The signal processing unit is configured to: when the health status of the current sampling period is normal, generate position output with the optical absolute position weight greater than the magnetic compensation position weight; when the health status of the current sampling period is optical degradation, gradually decrease the optical absolute position weight and gradually increase the magnetic compensation position weight according to a preset time sequence; when the health status of the current sampling period is magnetic interference, generate position output using only the optical absolute position.

8. The absolute encoder according to claim 7, characterized in that, The signal processing unit is configured to: when the health status of the current sampling period switches from a non-dual-path abnormal state to a dual-path abnormal state for the first time, record the last position output before the switch as the frozen position, and repeatedly output the frozen position during the period of maintaining the dual-path abnormal state; when no valid frozen position is obtained, output a fixed safe position according to the pre-stored safe position parameters, and output the health status information representing the dual-path abnormal state through the communication interface.

9. The absolute encoder according to claim 6, characterized in that, The signal processing unit is configured to: set a health status field in the data output through the communication interface, periodically update the current health status in the health status field, and simultaneously count the cumulative number of sampling cycles of optical degradation status and magnetic interference status, and output the current health status and cumulative occurrence count through the communication interface for the host system to make maintenance decisions.

10. The absolute encoder according to claim 1, characterized in that, The optical detection chip in the optical detection subsystem and the magnetic detection chip in the magnetic detection subsystem are mounted on the same circuit board. The circuit board is equipped with a dedicated integrated circuit chip, which integrates the optical detection circuit and the magnetic detection circuit. The relative positions of the optical sensitive area and the magnetic sensitive area are defined by the circuit board wiring and the layout design of the dedicated integrated circuit chip. The inherent alignment of the optical code disk and the magnetic ring is achieved by the cooperation of the package shell and the mechanical base.

Images