Linear motor motion position dynamic control method based on multi-sensor fusion

Abstract

The application discloses a kind of linear motor motion position dynamic control methods based on multi-sensor fusion, it is related to motor control technical field.The method collects the position reading and arrangement position of each reading head on mover, establishes the corresponding relationship of reading difference and reading head spacing;The temperature along the guide rail is measured and the thermal elongation at each place is determined, and the thermal expansion distribution is formed;With fixed reference starting point, the thermal drift amount difference of each reading head is converted, compared with the measured reading difference under the same spacing, and the false reading difference caused by heat is separated;Subtract the thermal drift amount with the change of temperature and dynamically update the position fusion starting line;Based on the starting line fusion, the position reading and reading difference of each reading head are calculated, and the real-time translational position and deflection attitude of mover are obtained.The application makes the attitude resolution gain of reading head spacing and the amplification of heat-induced error decoupled, improves the stability and accuracy of position and deflection attitude output.

Description

Technical Field

[0001] This invention relates to the field of motor control technology, and in particular to a method for dynamic control of the motion position of a linear motor based on multi-sensor fusion. Background Technology

[0002] Linear motors eliminate intermediate transmission links, directly driving the mover to complete reciprocating motion. They are commonly used in applications requiring high positioning accuracy, and the mover position is often obtained using an optical encoder arranged along the direction of motion. To obtain the mover's yaw attitude about a point perpendicular to the direction of motion beyond the translational position, multiple reading heads are typically placed on the mover at intervals along the direction of motion. The yaw angle information is obtained by comparing the readings from each reading head. Geometrically, the same yaw deflection results in a greater reading difference between two reading heads that are far apart. Therefore, increasing the baseline spacing between the reading heads helps improve the resolution of the yaw angle. The conventional practice based on this is to maximize the spacing between the reading heads along the direction of motion, assuming that a larger spacing yields a more reliable yaw attitude.

[0003] However, this judgment relies on the premise that the measurement reference remains thermally stable, which is often difficult to achieve in actual systems. Motor winding losses and guide rail friction continuously generate heat during operation, resulting in an uneven temperature distribution along the guide rail and constant changes in operating conditions and travel position. As the guide rail and mounting base expand due to heat, the position reference represented by the grating lines drifts accordingly. The cumulative expansion at each reading head's location differs, leading to varying drifts relative to the reference. If the drifts of each reading head were consistent, they could cancel each other out during reading comparison. However, the uneven temperature along the travel path causes a significant difference in drift, which cannot be eliminated during comparison. The larger the distance between reading heads, the more significant the temperature rise difference and cumulative expansion difference between the two locations, increasing the drift difference between reading heads and directly impacting the reading comparison process. This manifests as false sway angles and positional deviations that are difficult to distinguish from actual motion. It is evident that while increasing the spacing improves geometric resolution, it also amplifies thermally induced errors; these two effects are coupled through the spacing. When the thermally induced error exceeds the resolution gain, the accuracy of position and yaw attitude determination decreases rather than increases, making it difficult to maintain optimality under various temperature conditions using a fixed spacing. Furthermore, since the temperature distribution changes in real time with operating conditions, relying on pre-calibrated static corrections cannot track and eliminate this temperature-dependent drift. Therefore, how to stably obtain the position and yaw attitude of the mover under the condition that the reading head spacing simultaneously determines both the yaw angle resolution and the thermally induced drift error, and that the guide rail temperature distribution changes dynamically, has become a major technical problem to be solved. Summary of the Invention

[0004] To address the issues of simultaneously increasing the yaw angle resolution and thermal drift error due to the spacing between reading heads, and the difficulty in stably obtaining the mover position and yaw attitude when the guide rail temperature dynamically changes, this invention provides a dynamic control method for the motion position of a linear motor based on multi-sensor fusion. This method includes the following steps: S1. Collect the position readings of each reading head on the mover and their arrangement along the direction of motion, and analyze and process them to obtain the correspondence between the reading difference and the distance between the reading heads; S2. Measure the temperature along the guide rail according to the arrangement of each reading head, determine the thermal expansion at each point of the guide rail, and obtain the thermal expansion distribution along the guide rail. S3. Using a fixed reference starting point as a benchmark, calculate the thermal drift in the readings of each reading head based on the thermal expansion distribution along the guide rail, and obtain the difference in drift between the reading heads. S4. Compare the difference in thermal drift between reading heads at the same spacing with the measured reading difference to separate the reading difference component caused by the thermal expansion of the guide rail rather than the actual sway of the mover, and obtain the thermally induced false reading difference. S5. As the temperature along the guide rail changes in real time, the thermal drift in the readings of each reading head is deducted and corrected from the fixed reference starting point based on the thermally induced false reading difference, so as to obtain the position fusion starting line that is dynamically updated with temperature. S6. Based on the position fusion starting line that is dynamically updated with temperature, the position readings and reading differences of each reading head are fused and calculated to obtain the real-time translational position and yaw attitude of the mover after eliminating the influence of thermally induced false reading differences.

[0005] Compared with existing technologies, this invention establishes the thermal expansion distribution by measuring the temperature along the guide rail, separates the thermal drift and its differences at each reading head position from the measured reading difference, and dynamically updates the position fusion starting line with temperature changes. This decouples the gain of the reading head spacing on attitude resolution from the amplification of thermally induced errors, thereby improving the stability and accuracy of the real-time translational position and yaw attitude calculation results of the mover. Attached Figure Description

[0006] Figure 1 This is a schematic diagram of the overall process of the linear motor motion position dynamic control method based on multi-sensor fusion according to the present invention.

[0007] Figure 2 This is a schematic diagram of the thermally induced false reading difference determination and separation process of the present invention.

[0008] Figure 3 This is a schematic diagram of the dynamic update and consistency verification process of the position fusion starting line of the present invention. Detailed Implementation

[0009] The specific embodiments of the present invention will now be described with reference to the accompanying drawings.

[0010] It should be understood that the following embodiments are used to illustrate the present invention and are not intended to limit the scope of protection of the present invention.

[0011] In this embodiment, multiple grating ruler reading heads are arranged on the linear motor mover along the direction of motion, and temperature detection points are set up along the entire stroke of the mover. The motion controller synchronously collects position readings and guide rail temperatures, and sequentially completes the determination of thermal expansion distribution, separation of thermally induced false reading differences, dynamic calculation line update, and fusion calculation of translational position and yaw attitude.

[0012] S1. Collect the position readings of each reading head on the mover and their arrangement along the direction of motion, and analyze and process them to obtain the correspondence between the reading difference and the distance between the reading heads.

[0013] The idea behind this step is to organize the position readings of each reading head into reading differences according to the reading head pair, and bind them to the corresponding spacing, so that the attitude information is displayed in a form that is magnified with the spacing.

[0014] In a scenario where the linear motor's mover reciprocates along the guide rail, once the mover yaws around an axis perpendicular to the direction of motion, the same yaw angle will cause a greater reading difference between two reading heads that are far apart. The reading difference increases approximately linearly with the distance between the reading heads. Therefore, the correspondence between the reading difference and the distance contains both the mover's attitude information and the thermally induced error information to be separated later.

[0015] Specifically, the reading heads are first sorted according to their positions along the direction of movement to obtain the reading head arrangement order; then, the difference between the position readings of any two reading heads in the arrangement order is calculated to obtain the reading difference of the corresponding reading head pair; subsequently, the difference between the arrangement positions of the corresponding reading head pairs along the direction of movement is calculated to obtain the spacing of the corresponding reading head pairs; finally, the reading difference of each reading head pair is correlated with the corresponding spacing to obtain the correspondence between the reading difference and the reading head spacing.

[0016] In one embodiment, N reading heads are arranged along the direction of motion on the mover. N is typically 3, set based on the premise that at least two reading heads are needed to form a reading difference, and the third reading head provides a redundant reading head pair for cross-checking. N can be adjusted within the range of 2 to 6 depending on the mover length and cost constraints. The spacing between adjacent reading heads is typically 200 mm, set based on balancing yaw angle resolution and structural compactness within the installation space allowed by the mover structure, and can be adjusted within the range of 50 to 600 mm. The resolution of the grating ruler is typically 0.1 μm, set based on the requirement of sub-micron position feedback for precision linear motor positioning, and can be selected and adjusted within the range of 0.05 to 0.5 μm. The position reading sampling frequency is typically 1 kHz, set based on matching the position loop control cycle and covering the position change rate at the mover's highest speed, and can be adjusted within the range of 500 Hz to 10 kHz.

[0017] Furthermore, the arrangement position of the reading heads refers to the position coordinates of their measurement reference points along the direction of motion, which changes in real time with the movement of the mover. The difference between the arrangement positions of any two reading heads, i.e., the spacing between the reading head pairs, is determined by the relative installation relationship on the mover and does not change with the movement of the mover. It serves as the geometric amplification baseline for the reading difference; the larger the spacing, the larger the reading difference mapped to the same yaw angle. The reading difference includes both the geometric component caused by the actual yaw of the mover and the thermal component to be separated in subsequent steps. Position readings and reading differences are expressed in μm, and spacing is expressed in mm. The position readings of each reading head are synchronously latched by the position acquisition interface of the motion controller according to the position sampling frequency. The calculation of the reading difference and the calculation of subsequent steps are performed by the processing unit of the motion controller.

[0018] It should be noted that the correspondence is organized into a spacing-reading difference data set with the reading head pair as the index, which is used by step S4 for spacing alignment and comparison and step S6 for calculating the sway angle.

[0019] S2. Based on the arrangement of the reading heads collected in step S1, the temperature along the guide rail is measured to determine the thermal elongation at various points on the guide rail, thus obtaining the thermal expansion distribution along the guide rail. The idea behind this step is to replace the single-point temperature assumption with segmented temperature measurements along the guide rail, transforming the temperature distribution into a quantitative description of the reference drift at various points on the guide rail.

[0020] In operating scenarios where motor winding losses and guide rail friction continuously generate heat, the temperature distribution along the guide rail is not uniform and constantly changes with operating conditions and travel position. As the guide rail and mounting base expand due to heat, the position reference represented by the grating lines drifts, and this drift varies from place to place. Only by segmenting the temperature rise along the travel path by position can the different reference drifts at each reading head be characterized, providing a basis for subsequent conversions.

[0021] Specifically, temperature detection points are set up at the locations of each reading head along the movement of the mover, and the temperature values ​​at each detection point are collected to obtain the temperature distribution along the guide rail. Based on the temperature change of each detection point relative to the reference temperature, and combined with the thermal expansion characteristics of the guide rail material, the thermal expansion at each point of the guide rail is determined. The thermal expansion at each point is accumulated along the direction of motion to obtain the thermal expansion distribution along the guide rail. Since the mover reciprocates within its stroke, the locations of each reading head traverse the entire stroke. Therefore, temperature detection points are set up along the entire stroke of the mover, and the interval between detection points along the direction of motion is no greater than the distance between adjacent reading heads, so that there is a detection point near the location of each reading head at any given time.

[0022] For example, a platinum resistance temperature sensor is used for the temperature detection point. The typical temperature measurement accuracy is ±0.1℃. This is based on the fact that the ±0.1℃ temperature measurement uncertainty corresponds to an elongation uncertainty of approximately 0.12 to 0.24 μm in a 100–200 mm detection segment, which is on the same order of magnitude as the submicron resolution of the grating and consistent with the manufacturer's nominal temperature measurement accuracy. The temperature measurement accuracy can be adjusted within the range of ±0.05 to ±0.5℃. The typical reference temperature is 20℃, set according to the reference temperature of the geometric measurement standard specified in ISO 1, and can be adjusted within the range of 15–25℃ according to the ambient temperature during equipment calibration. The typical temperature sampling period is 100 ms, set based on the fact that the time constant of the guide rail thermal process is on the order of tens of seconds to minutes. This period is sufficient to track temperature changes without distortion and the data processing volume is negligible, and can be adjusted within the range of 10 ms to 1 s.

[0023] Furthermore, thermal expansion conversion is performed based on the temperature sequence of each detection point output by the temperature acquisition module according to the temperature sampling period. The guide rail segment between adjacent temperature detection points is used as the segment unit, with the length l of the k-th segment. k Take the distance between adjacent detection points, and the temperature rise ΔT. k Take the average of the measured temperature rise (the change in measured temperature relative to the reference temperature) at both ends of the section. The formula for calculating the thermal elongation of the k-th section of the guide rail is: Where α is the coefficient of linear expansion of the guide rail material, in °C. -1 ;ΔT k The unit is ℃; l k The unit is mm; δ k This represents the thermal elongation of this section, calculated and converted to μm; its physical meaning is the increase in length of this section of the guide rail due to temperature rise. α is typically taken as 11.8 × 10⁻⁶. -6 ℃ -1 (For steel guide rails), the setting is based on the coefficient of linear expansion at room temperature given in the material manual, and can be adjusted according to the guide rail material within 5×10. -6 ℃ -1 (Granite) up to 23×10 -6 ℃ -1 The value is taken within the range of (aluminum alloy). The formula for calculating the thermal expansion distribution along the guide rail is: .

[0024] The summation range includes all segments whose position coordinates along the direction of motion are not greater than p. Here, p is the position coordinate along the direction of motion, in mm; E(p) is the cumulative thermal expansion from the starting end of the guide rail to position p, in μm; its physical meaning is the displacement of the reference point at position p relative to the starting end due to thermal expansion.

[0025] Preferably, when the distance between adjacent reading heads is large, a temperature detection point is added between the two reading head locations. The typical distance between the detection points is 100mm. The basis for setting this distance is to ensure that the temperature difference between adjacent detection points does not exceed the order of magnitude corresponding to the temperature measurement accuracy and to ensure the continuity of the distribution characterization along the process. It can be adjusted within the range of 50 to 300mm.

[0026] It is understandable that the temperature field changes much slower than the motion of the mover. Temperature sampling and position sampling can be performed at different rates, and they can be aligned by timestamps before proceeding to subsequent steps.

[0027] S3. Using a fixed reference starting point as a benchmark, calculate the thermal drift in the readings of each reading head based on the thermal expansion distribution along the guide rail obtained in step S2, and determine the difference in drift between reading heads. The idea behind this step is to map the thermal expansion distribution along the guide rail to the position of each reading head, obtain the thermal drift of each reading head, and then convert it into a difference expression between reading head pairs.

[0028] In a scenario where one end of the guide rail is mechanically anchored and thermal expansion accumulates unidirectionally towards the free end, the reading head farther from the fixed end experiences a greater cumulative expansion, and the drift of each reading head relative to the reference is different. If the drift of each reading head is consistent, they can cancel each other out during reading comparison; however, uneven temperature along the guide rail causes a significant difference in drift, which cannot be eliminated during comparison and is precisely what needs to be quantitatively extracted in this step.

[0029] Specifically, the distance between the arrangement position of each reading head and the fixed reference starting point along the direction of motion is determined to obtain the reference distance of each reading head. Based on the thermal expansion distribution along the guide rail, the thermal expansion within the reference distance interval of each reading head is accumulated to obtain the cumulative thermal expansion at each reading head position. The cumulative thermal expansion at each reading head position is converted into the offset in the corresponding position reading to obtain the thermal drift in the reading of each reading head. The thermal drift in the reading of the reading head is the value of the thermal expansion distribution at the reference distance of the reading head, expressed in μm. Its physical meaning is the offset component introduced by thermal expansion in the reading at the reading head position, which is unrelated to the actual motion of the mover.

[0030] In one possible implementation, the difference in thermal drift between the readings of two adjacent reading heads is calculated to obtain the drift difference between adjacent reading heads; the distance between two adjacent reading heads along the direction of movement is obtained, and the drift difference is correlated with the corresponding distance to obtain the relationship between the drift difference and the distance; based on this relationship, the drift differences of adjacent reading head pairs are accumulated segment by segment along the arrangement order to determine the thermal drift difference of any reading head pair, thus obtaining the thermal drift difference between reading heads. The difference in thermal drift between the i-th and j-th reading heads is calculated to obtain the thermal drift difference Δε between reading heads. ij Its approximate calculation formula is: .

[0031] Where, ΔT ij L represents the average temperature rise of the guide rail section between the i-th and j-th reading heads, in °C. ij The distance between the reading heads is in mm; Δε ij The difference in thermal drift is calculated and converted to μm. Its physical meaning is that thermal expansion injects a spurious component into the reading difference of the reading head, and it increases approximately linearly with the spacing. This is a quantitative expression of the coupling relationship between the reading head spacing and the amplification of thermally induced error while improving geometric resolution.

[0032] For example, when the average temperature rise between the two reading heads is 1°C, the spacing is 200 mm, and the guide rail is made of steel, the difference in thermal drift is about 2.4 μm, which is far greater than the grating resolution of 0.1 μm. It can be seen that this spurious component must be explicitly extracted and subtracted.

[0033] Furthermore, the fixed reference starting point is preferably located at the anchoring end where the guide rail and base are rigidly connected, or at the zero mark of the grating ruler. This is based on the fact that the expansion displacement at this location is mechanically constrained to approximately zero, thus preventing reference drift. The reference starting point can be changed to any anchoring position where expansion is constrained, depending on the actual installation configuration. Since the reading head moves with the actuator, the reference distance of each reading head is determined relative to the fixed reference starting point based on its current position reading, and the actuator position is updated in real time. The guide rail segment can be determined using the position reading before correction; the micrometer-level correction has negligible impact on segment assignment.

[0034] It should be noted that the thermal expansion of the location between temperature detection points is obtained by linear interpolation between adjacent detection points; the basis for using linear interpolation is that the temperature gradient between detection points is gentle and the error introduced by linear approximation is less than the elongation corresponding to the temperature measurement accuracy.

[0035] S4. Compare the difference in thermal drift between reading heads obtained in step S3 at the same spacing with the difference in measured readings obtained in step S1, and separate the reading difference component caused by the thermal expansion of the guide rail rather than the actual sway of the mover, thus obtaining the thermally induced spurious reading difference. The idea of ​​this step is to use the difference in thermal drift calculated independently from temperature information as a reference to identify and remove the thermally induced component from the measured reading difference.

[0036] In operating scenarios where both actual yaw and thermal drift of the mover coexist, the difference in measured readings includes both a geometric component that increases linearly with the spacing and a thermal component that also increases with the spacing. These two effects are coupled together through the spacing and cannot be distinguished by the difference in readings alone. The difference in thermal drift, which is independently calculated from temperature information, must be used as a reference to separate the two.

[0037] Specifically, the measured reading difference and the thermal drift difference between the reading heads are aligned according to the reading head spacing to obtain comparison data pairs at the same spacing. At the same spacing, the thermal drift difference between the reading heads and the measured reading difference are compared to obtain the deviation between them. Based on the deviation, it is determined whether the measured reading difference contains a component exceeding the true sway, thus obtaining the reading difference comparison result. The deviation, expressed in μm, is the difference between the measured reading difference and the thermal drift difference between the reading heads at the same spacing. Its physical meaning is the estimated reading difference attributable to the true sway of the mover after deducting the thermally induced reference from the measured reading difference. During the determination, the measured reading difference is compared with the deviation; the difference corresponds to the thermal drift difference. When the absolute value of the thermal drift difference between the reading heads is greater than the determination threshold, it is determined that the measured reading difference contains a thermally induced false reading difference caused by the thermal expansion of the guide rail, forming the reading difference comparison result.

[0038] The typical threshold is 0.5μm. It is set based on the repeatability index given by the grating ruler manufacturer and the reading noise 3σ level calibrated by the statistical data from the static sampling test on site. The thermal component smaller than this value is submerged in the noise and does not need to be deducted. The threshold can be adjusted in the range of 0.2 to 2μm according to the on-site noise level.

[0039] For example, a component corresponding to the difference in thermal drift between reading heads is extracted from the measured reading difference to obtain the thermal expansion-related component; the remaining portion after deducting the thermal expansion-related component from the measured reading difference is marked as the component generated by the actual yaw of the mover to obtain the actual yaw component; the thermal expansion-related component is determined as the reading difference component caused by the thermal expansion of the guide rail, i.e., the thermally induced spurious reading difference. The thermally induced spurious reading difference is a spurious component that appears in the reading difference even if the mover does not yaw at all; the actual yaw component is a geometric component caused only by the actual deflection of the mover around an axis perpendicular to the direction of motion, independent of the temperature state, and both are measured in μm.

[0040] Preferably, a moving average filter is applied to the difference in measured readings before comparison. The window length is typically 10 sampling points. The setting is based on forming a 10ms smooth window at a 1kHz sampling position, which suppresses random noise while fully preserving the minute-level slow change process of thermal drift. The window length can be adjusted within the range of 5 to 50 sampling points.

[0041] It should be noted that when there are three or more reading heads, the comparison is performed in parallel on multiple reading head pairs, and the comparison results of redundant reading head pairs can be used to cross-validate the consistency of the separation results.

[0042] S5. As the temperature along the guide rail changes in real time, based on the thermally induced spurious reading difference obtained in step S4, the thermal drift in the readings of each reading head is subtracted and corrected from the fixed reference starting point to obtain the position fusion starting line that is dynamically updated with temperature. The idea behind this step is to make the subtraction action and the starting reference linked in real time with the guide rail temperature, transforming static correction into dynamic correction that updates with temperature.

[0043] In scenarios where operating conditions and travel positions are constantly changing and the guide rail temperature field is continuously evolving, thermal drift is not a constant that can be fixed by a single calibration, but a process quantity that changes in real time with temperature. Relying on pre-calibrated static corrections cannot track and eliminate this temperature-dependent drift deviation. Only by making the starting point for position fusion updated synchronously with the temperature rise can the subtraction action always correspond to the current thermal state.

[0044] Specifically, the temperature change along the guide rail is acquired in real time, and the thermal expansion distribution along the guide rail is updated according to the change. Based on the updated thermal expansion distribution and the difference between thermally induced spurious readings, the current thermal drift correction amount of each reading head is determined from a fixed reference starting point. The thermal drift amount in the readings of each reading head is deducted once according to the thermal drift correction amount to obtain the corrected readings of each reading head. The starting line used for fusion is updated according to the corrected readings of each reading head to obtain the position fusion starting line that is dynamically updated with temperature.

[0045] Furthermore, when the reading difference comparison result in step S4 indicates a deviation between the measured thermally induced spurious reading difference and the thermal drift difference between the temperature-converted reading heads, the processing unit corrects the correction amount obtained from the temperature conversion according to the deviation of each reading head pair: taking the reading head with the smallest deviation as the anchor point, the measured thermally induced spurious reading differences of the remaining reading head pairs are accumulated segment by segment to the anchor point correction amount to obtain the thermal drift correction amount of each reading head after actual measurement correction; when the comparison result is consistent with the temperature conversion, the cumulative value of the thermal expansion distribution is directly used as the correction amount. The current thermal drift correction amount of each reading head is taken as the cumulative thermal elongation of the guide rail along the reference distance of the reading head after the update; the position reading of the reading head is subtracted from the correction amount to obtain the corrected reading head reading, which is equivalent to the position reading returning to the reference temperature state; at the same time, the starting zero position of each reading head is moved forward synchronously by the same correction amount, that is, the update of the position fusion starting line is completed, and the reading deduction and the reference update are equivalent. After the above deductions, the thermally induced spurious reading difference in the difference between the corrected readings of the same reading head is eliminated. The correction amount, reading, and starting line values ​​are all expressed in μm.

[0046] In one embodiment, the starting line update cycle is typically set to 100ms and synchronized with the temperature sampling cycle. The basis for this setting is that the thermal process time constant is much larger than this cycle, enabling cycle-by-cycle tracking without hysteresis and the additional computational load of the position loop is negligible. The update cycle can be adjusted within the range of 50ms to 1s.

[0047] Furthermore, a temperature change trigger threshold is set. The starting line update is only performed when the temperature change of any temperature detection point relative to the last update exceeds the threshold. The threshold is typically 0.1℃, and the setting is based on the temperature measurement accuracy of the selected platinum resistance temperature sensor (consistent with step S2, ±0.1℃). Changes smaller than this value do not produce a distinguishable drift increment. The threshold can be adjusted within the range of 0.05 to 0.5℃.

[0048] It is understood that the one-time deduction means that each update cycle makes a full correction to the reading according to the correction amount at the current moment, rather than incrementally adding the correction result of the previous cycle, thereby avoiding the accumulation of incremental error with the number of updates and ensuring that the starting line does not experience correction drift during long-term operation.

[0049] S6. Based on the dynamically updated position fusion starting line obtained in step S5, the position readings and reading differences of each reading head are fused and calculated to obtain the real-time translational position and yaw attitude results of the mover after eliminating the influence of thermally induced spurious reading differences. The idea of ​​this step is to complete the fusion calculation of translational position and yaw attitude on the dynamic starting line, and to ensure the validity and continuity of the output with redundancy consistency verification.

[0050] In scenarios where thermally induced components have been subtracted cycle by cycle, the corrected reading difference retains the true geometric components that are linearly amplified with the spacing. In this case, increasing the reading head spacing only improves the angular resolution: the gain of spacing on resolution and the amplification of thermally induced errors are decoupled through the dynamic reference. Large-spacing reading head pairs can leverage the advantages of attitude calculation without introducing thermally induced deviations, and the mover position curve is no longer disturbed by spurious yaws, resulting in a more stable output.

[0051] Specifically, using the dynamically updated position fusion reference line based on temperature as a baseline, the position readings of each reading head are reduced to obtain the corrected position readings for each reading head; the corrected position readings are then combined to obtain the real-time translational position of the mover; the differences between each pair of corrected position readings are calculated to obtain the corrected reading difference; based on the correspondence between the corrected reading difference and the distance between the reading heads, the yaw angle of the mover is calculated to obtain the yaw attitude of the mover; the real-time translational position and yaw attitude are then combined and output. The formula for calculating the real-time translational position is: The weights satisfy Σw i =1. Where, x' i (t) represents the corrected position reading of the i-th reading head obtained in step S5, in μm; w i The fusion weight is dimensionless; X(t) is the real-time translational position of the mover, in μm; its physical meaning is the weighted sum of the corrected readings of each reading head, representing the overall translational position of the mover's measurement section.

[0052] wi Typically, equal weighting is used, i.e., 1 / N. The basis for this setting is that all reading heads are of the same model and have the same noise level. i The signal-to-noise ratio can be adjusted and normalized based on the measured signal-to-noise ratio of each reading head, with each weight fluctuating by no more than 50% above or below 1 / N. The approximate formula for calculating the yaw angle is: Wherein, Δx' ij The corrected reading difference, i.e., the difference between the corrected readings of the two reading heads, is equal in value to the true yaw component separated in step S4 at the same moment, and its unit is μm; L ij This refers to the spacing between corresponding reading head pairs, in mm, and is calculated in conjunction with Δx'. ij All are converted to the same unit of length; θ(t) is the yaw angle of the mover, in rad; its physical meaning is the actual deflection angle of the mover around the axis perpendicular to the direction of motion. The basis for the small angle approximation is that the yaw angle is usually in the range of arcseconds to angular components.

[0053] Preferably, the reading head pair with the largest spacing is selected first when calculating the yaw angle. The basis for this setting is that after eliminating thermally induced spurious reading differences, the largest spacing provides the highest angular resolution. The selection of reading head pairs can be adjusted among the various reading head pairs, or the calculation results of multiple pairs can be weighted and combined according to the spacing.

[0054] Furthermore, this step also includes a consistency verification and maintenance strategy for the solution results. The consistency of the yaw angles calculated separately for redundant reading head pairs is verified, and the maximum difference between each pair of calculated values ​​is taken as the consistency residual. This residual directly reflects the accuracy of the thermal expansion distribution representation at the solution level; the more thorough the thermally induced component subtraction, the more consistent the calculated values. A typical residual threshold is 1.5 × 10⁻⁶. -6 The value is rad (approximately 0.3 arcseconds). This value was determined through statistical measurements of the dispersion of the calculated yaw angle values ​​of each reading head under constant temperature conditions during the assembly and debugging phase. It is approximately three times the fluctuation range of 0.1μm reading noise propagating through a 200mm baseline, and the residual threshold can be around 1×10⁻⁶. -6 ~1×10 -5 Adjust within the rad range.

[0055] When the consistency residual is not greater than the residual threshold, the calculation is deemed valid, and the real-time translational position and yaw attitude are output normally. When the consistency residual is greater than the residual threshold, the thermal expansion distribution update is deemed to be lagging or the temperature detection is abnormal. At this time, the most recently verified position fusion starting line is maintained, the starting line update is frozen, and the calculation continues on the starting line according to the position sampling period and the real-time translational position and yaw attitude are output as a protective process. At the same time, the return step S2 is triggered to remeasure the temperature along the guide rail. The starting line is updated again through steps S3 to S5. After the consistency verification is passed, the normal calculation output is restored.

[0056] It should be noted that the reduction and solution are performed cycle by cycle according to the position sampling period, and the starting line is refreshed only according to the update cycle of step S5. The two-level cycle decoupling not only ensures the real-time position feedback, but also makes the output real-time position curve remain stable and continuous under the condition of continuous change of guide rail temperature.

[0057] It is understandable that by cyclically executing steps S2 to S6 during the update cycle, a complete closed loop is formed to stably determine the position and yaw attitude of the mover under temperature drift conditions.

Claims

1. A method for dynamic control of the motion position of a linear motor based on multi-sensor fusion, characterized in that, include: S1. Collect the position readings of each reading head on the mover and their arrangement along the direction of motion, and analyze and process them to obtain the correspondence between the reading difference and the distance between the reading heads; S2. Measure the temperature along the guide rail according to the arrangement of each reading head, determine the thermal expansion at each point of the guide rail, and obtain the thermal expansion distribution along the guide rail. S3. Using a fixed reference starting point as a benchmark, calculate the thermal drift in the readings of each reading head based on the thermal expansion distribution along the guide rail, and obtain the difference in drift between the reading heads. S4. Compare the difference in thermal drift between reading heads at the same spacing with the measured reading difference to separate the reading difference component caused by the thermal expansion of the guide rail rather than the actual sway of the mover, and obtain the thermally induced false reading difference. S5. As the temperature along the guide rail changes in real time, the thermal drift in the readings of each reading head is deducted and corrected from the fixed reference starting point based on the thermally induced false reading difference, so as to obtain the position fusion starting line that is dynamically updated with temperature. S6. Based on the position fusion starting line that is dynamically updated with temperature, the position readings and reading differences of each reading head are fused and calculated to obtain the real-time translational position and yaw attitude of the mover after eliminating the influence of thermally induced false reading differences.

2. The method according to claim 1, characterized in that, The analysis and processing yields the correspondence between the reading difference and the reading head spacing, specifically including: The reading heads are sorted according to their position along the direction of movement to obtain the reading head arrangement order; The difference between the position readings of any two reading heads in the arrangement order is obtained to obtain the reading difference of the corresponding reading head pair; Calculate the difference in the arrangement positions of the corresponding reading head pairs along the direction of movement to obtain the spacing between the corresponding reading head pairs; By associating the reading difference of each pair of reading heads with the corresponding spacing, the correspondence between the reading difference and the reading head spacing is obtained.

3. The method according to claim 1, characterized in that, The method of measuring the temperature along the guide rail based on the arrangement of each reading head, determining the thermal elongation at various points on the guide rail, and obtaining the thermal expansion distribution along the guide rail specifically includes: Temperature detection points are set up at the locations of each reading head along the travel of the mover, and the temperature values ​​of each temperature detection point are collected to obtain the temperature distribution along the guide rail. Based on the temperature change of each temperature detection point relative to the reference temperature, and combined with the thermal elongation characteristics of the guide rail material, the thermal elongation at each point on the guide rail is determined. The thermal expansion distribution along the guide rail is obtained by summing up the thermal elongation at each location along the direction of motion.

4. The method according to claim 1, characterized in that, The calculation of thermal drift in the readings of each reading head, based on a fixed reference starting point and the thermal expansion distribution along the guide rail, specifically includes: The distance between the fixed reference starting point and the location of each reading head along the direction of movement is determined to obtain the reference distance of each reading head; Based on the thermal expansion distribution along the guide rail, the thermal expansion within the reference distance interval of each reading head is accumulated to obtain the cumulative thermal expansion at each reading head position. The cumulative thermal expansion at each reading head position is converted into the offset at the corresponding position reading to obtain the thermal drift in each reading head.

5. The method according to claim 1, characterized in that, The determination of the drift difference between reading heads, specifically the thermal drift difference between reading heads, includes: The difference in thermal drift between two adjacent reading heads is obtained by calculating the difference in drift between them. The distance between two adjacent reading heads along the direction of movement is obtained, and the drift difference is correlated with the corresponding distance to obtain the relationship between the drift difference and the distance. Based on the relationship between the drift difference and the spacing, the thermal drift difference of each reading head pair is determined, and the thermal drift difference between reading heads is obtained.

6. The method according to claim 1, characterized in that, The comparison of the thermal drift difference between reading heads at the same spacing with the measured reading difference specifically includes: Align the measured reading difference with the thermal drift difference between the reading heads according to the reading head spacing to obtain comparison data pairs under the same spacing. By comparing the difference in thermal drift between reading heads with the difference in measured readings at the same spacing, the deviation between the two can be obtained. Based on the deviation, determine whether the measured reading difference contains a component that exceeds the true swing, and obtain the reading difference comparison result.

7. The method according to claim 1, characterized in that, The separation of the reading difference component caused by the thermal expansion of the guide rail, rather than the actual wobble of the mover, to obtain the thermally induced spurious reading difference specifically includes: The thermal expansion correlation component is obtained by extracting the component corresponding to the difference in thermal drift between the reading heads from the measured reading difference. The remaining part after deducting the thermal expansion-related component from the measured reading difference is marked as the component generated by the true yaw of the mover, thus obtaining the true yaw component. The thermal expansion correlation component is determined as the reading difference component caused by the thermal expansion of the guide rail, thus obtaining the thermally induced spurious reading difference.

8. The method according to claim 1, characterized in that, The real-time temperature change along the guide rail, based on the thermally induced spurious reading difference, is used to subtract and correct the thermal drift in the readings of each reading head from a fixed reference starting point, resulting in a dynamically updated position fusion starting line that updates with temperature. Specifically, this includes: Real-time acquisition of temperature changes along the guide rail, and updating of thermal expansion distribution along the guide rail based on the changes; Based on the updated thermal expansion distribution and the difference in thermally induced spurious readings, the current thermal drift correction for each reading head is determined from a fixed reference starting point; The thermal drift in each reading head is subtracted once according to the thermal drift correction amount to obtain the corrected readings of each reading head. The starting line for fusion is updated based on the corrected readings of each reading head, resulting in a position fusion starting line that is dynamically updated with temperature.

9. The method according to claim 1, characterized in that, The process of fusing and calculating the readings and differences at each reading head position to obtain the real-time translational position and yaw attitude of the mover after eliminating the influence of thermally induced spurious reading differences specifically includes: Using the position fusion starting line that is dynamically updated with temperature as a reference, the position readings of each reading head are calculated to obtain the corrected position readings of each reading head; By combining the readings at each corrected position, the real-time translational position of the mover is obtained. The difference between the readings at each correction position is obtained to get the corrected reading difference. The yaw angle of the mover is calculated based on the correspondence between the corrected reading difference and the distance between the reading heads, and the yaw attitude of the mover is obtained. The real-time translational position and yaw attitude are combined and output to obtain the real-time translational position and yaw attitude of the mover after eliminating the influence of thermally induced spurious reading differences.

Images