Patch machine mounting pressure closed-loop control method and system based on multi-sensor fusion

Abstract

The present application relates to the technical field of electronic assembly, and discloses a patch machine mounting pressure closed-loop control method and system based on multi-sensor fusion, which comprises the following steps: obtaining the vacuum pressure base frequency and timing real-time speed of the mounting head in the target patch machine, performing motion state synchronous cutting on the vacuum pressure base frequency to obtain a vacuum pressure sub-sequence and a vacuum pressure baseline value; performing normalized pressure drop amplitude analysis on the vacuum pressure sub-sequence to obtain an approximate pressure drop amplitude sequence; performing time-domain feature fitting on the approximate pressure drop amplitude sequence to obtain an approximate response time constant and an approximate response steady-state amplitude; performing gap height mapping on the approximate response steady-state amplitude to obtain a minimum gap height; performing aerodynamic impedance inversion on the element to obtain a lower surface attitude deflection angle; and performing contact attitude pre-compensation on the mounting head to obtain mounting stroke parameters; and the present application can improve the efficiency of mounting pressure pre-control.

Description

Technical Field

[0001] This invention relates to the field of electronic assembly technology, and in particular to a method and system for closed-loop control of placement pressure in a pick-and-place machine based on multi-sensor fusion. Background Technology

[0002] Pick and place machines are core equipment for electronic assembly of circuit boards. The placement head relies on vacuum adsorption to complete the substrate placement operation. The dynamic response of vacuum pressure, the gap between the component and the substrate, and the orientation of the component bottom surface directly determine the placement pressure control accuracy. Existing equipment mostly uses vacuum pressure sensors and lifting shaft position sensors to independently collect working condition data, and only uses vacuum pressure as a detection signal for the presence or absence of components, without carrying out multi-sensor time sequence data fusion and calculation.

[0003] Existing technologies rely solely on single-position signals or single-point vacuum pressure to control the mounting stroke. They lack a mechanism for synchronizing vacuum pressure timing with lifting shaft speed, cannot filter out high-frequency pressure ripple interference generated by drive shaft movement, and cannot invert the real-time gap and bottom surface tilt state of components through pressure time-domain response characteristics. They only use fixed parameters to preset the mounting stroke, and cannot pre-compensate for contact deviations caused by component tilt and substrate height differences. The pressure fluctuation during the mounting process is large, which can easily lead to process defects such as component pressure damage, solder joint defects, and component misalignment. The accuracy and process stability of the mounting pressure closed-loop control are difficult to guarantee. Therefore, how to rely on multi-sensor fusion to analyze the dynamic characteristics of vacuum pressure, invert the component gap and attitude, and achieve pre-closed-loop control of mounting pressure has become an urgent technical problem to be solved. Summary of the Invention

[0004] This invention provides a closed-loop control method and system for placement pressure of a chip mounter based on multi-sensor fusion, in order to solve the problems mentioned in the background art.

[0005] To achieve the above objectives, the present invention provides a closed-loop control method for placement pressure of a pick-and-place machine based on multi-sensor fusion, comprising:

[0006] S01. Obtain the vacuum pressure fundamental frequency and real-time speed of the placement head in the target placement machine, and synchronously cut the uniform descent segment of the vacuum pressure fundamental frequency to obtain the vacuum pressure subsequence and vacuum pressure baseline value of the placement head.

[0007] S02. Using the vacuum pressure baseline value as a reference, perform normalized pressure drop analysis on the vacuum pressure subsequence to obtain the approximate pressure drop sequence of the mounting head.

[0008] S03. Perform first-order inertial element time-domain feature fitting on the approach pressure drop sequence to obtain the approach response time constant and approach response steady-state amplitude of the mounting head;

[0009] S04. Map the proximity response steady-state amplitude to obtain the minimum gap height between the component carried in the mounting head and the target substrate.

[0010] S05. Based on the proximity response time constant and the proximity response steady-state amplitude, perform aerodynamic impedance inversion on the lower surface of the carrying element to obtain the attitude deflection angle of the lower surface of the carrying element.

[0011] S06. Based on the lower surface attitude skew angle and the minimum gap height, perform contact attitude pre-compensation on the mounting head to obtain the mounting stroke parameters of the mounting head.

[0012] In a preferred embodiment, obtaining the vacuum pressure fundamental frequency and real-time timing speed of the placement head in the target placement machine includes:

[0013] At the start of the current placement stroke, the pressure of the vacuum passage cavity of the placement head in the target pick and place machine is continuously sampled in the time domain to obtain the time-series vacuum pressure sequence of the placement head, and the position of the lifting drive shaft of the placement head is continuously sampled in the time domain to obtain the time-series position feedback sequence of the placement head.

[0014] Based on a preset notch frequency, the time-series vacuum pressure sequence is subjected to stopband filtering to eliminate the high-frequency drive ripple components superimposed in the time-series vacuum pressure sequence, while retaining the low-frequency trend components in the time-series vacuum pressure sequence, thereby obtaining the vacuum pressure fundamental frequency of the mounting head.

[0015] The displacement difference between adjacent sampling points in the time-series position feedback sequence is extracted, and the ratio between the displacement difference and the sampling time interval between the adjacent sampling points is used as the real-time speed of the mounting head.

[0016] In a preferred embodiment, the step of synchronously cutting the uniformly decreasing segment of the vacuum pressure fundamental frequency to obtain the vacuum pressure sub-sequence and vacuum pressure baseline value of the mounting head includes:

[0017] The starting and ending sampling points are calibrated to maintain a constant speed in the real-time timing sequence.

[0018] Based on the start sampling point time and the end sampling point time, all position sampling points are extracted from the time-series position feedback sequence, and the time span of all position sampling points is used as the uniform motion time window of the mounting head.

[0019] Based on the uniform motion time window, the continuous pressure waveform of the vacuum pressure fundamental frequency is cut out along the time axis to obtain the vacuum pressure sub-sequence of the mounting head;

[0020] Select a stable waveform in the vacuum pressure subsequence that has not yet shown a pressure drop trend, accumulate the pressure amplitude of the stable waveform, and use the quotient of the accumulated value and the total number of pressure amplitudes as the vacuum pressure baseline value of the mounting head.

[0021] In a preferred embodiment, the step of performing normalized pressure drop analysis on the vacuum pressure subsequence using the vacuum pressure baseline value as a reference to obtain the approximate pressure drop sequence of the mounting head includes:

[0022] The net pressure change of the vacuum pressure subsequence is obtained by subtracting the amplitude of the sampling points in the vacuum pressure subsequence from the baseline value of the vacuum pressure bit by bit.

[0023] Using the vacuum pressure baseline value as a reference, the net pressure change is divided by the vacuum pressure baseline value, and the quotients are arranged sequentially to obtain the proximity pressure drop sequence of the mounting head.

[0024] In a preferred embodiment, the step of performing first-order inertial element time-domain feature fitting on the proximity pressure drop sequence to obtain the proximity response time constant and proximity response steady-state amplitude of the mounting head includes:

[0025] By backtracking point by point from the starting point of the approaching pressure drop sequence, the tail convergence segment is obtained;

[0026] The pressure drop values ​​of all sampling points within the tail convergence section are accumulated and summarized, and the quotient between the accumulated summation result and the total number of sampling points within the tail convergence section is taken as the proximity response steady-state amplitude of the mounting head.

[0027] Starting from the beginning of the approach pressure drop sequence, the difference between the pressure drop value at the sampling point in the approach pressure drop sequence and the pressure drop value at the beginning point is statistically analyzed point by point along the positive direction of the time axis.

[0028] When the difference first reaches a specified percentage of the near-response steady-state amplitude, the characteristic sampling position is obtained;

[0029] The time span between the feature sampling position and the starting point is used as the proximity response time constant of the mounting head.

[0030] In a preferred embodiment, the step of backtracking point by point from the starting point of the approximate pressure drop sequence to obtain the tail convergence segment includes:

[0031] Starting from the beginning of the approach pressure drop sequence, traverse backward point by point to extract the variation between each sampling point in the approach pressure drop sequence and the previous sampling point.

[0032] Define a continuous segment where the change amount is first continuously lower than a preset tolerance limit, and mark the starting boundary of the continuous segment as the convergence start point;

[0033] All sampling points between the convergence start point and the termination point of the near pressure drop sequence are defined as the tail convergence segment.

[0034] In a preferred embodiment, mapping the proximity response steady-state amplitude to obtain the minimum gap height between the component carried in the mounting head and the target substrate includes:

[0035] Based on multiple sets of known gap heights, the vacuum pressure response of the target placement machine is measured to obtain the measured steady-state amplitude. The multiple sets of known gap heights and the measured steady-state amplitude are then compiled into a mapping record set.

[0036] Using the near-response steady-state amplitude as a matching benchmark, the mapping record set is traversed, and the absolute value of the difference between the near-response steady-state amplitude and the measured steady-state amplitude is calculated.

[0037] The minimum value is selected from the absolute values ​​of the differences, and the known gap height of the minimum value is used as the reference height.

[0038] Using the reference height as the positioning center, extract the upper neighbor height and lower neighbor height of the reference height from the mapping record set;

[0039] Based on the steady-state amplitude of the proximity response, the upper neighbor height and the lower neighbor height are linearly proportionally interpolated to obtain the minimum gap height.

[0040] In a preferred embodiment, the step of performing aerodynamic impedance inversion on the lower surface of the carrying element based on the proximity response time constant and the steady-state amplitude of the proximity response to obtain the attitude skew angle of the lower surface of the carrying element includes:

[0041] Obtain the outline dimensions of the component carried in the mounting head, and perform a half-section of the outline dimensions along the symmetrical midline of the component to obtain the half-size of the component.

[0042] Based on the proximity response time constant and the half-size, the lower surface attitude skew angle of the carrying element is calculated, wherein the formula for calculating the lower surface attitude skew angle is:

[0043] ;

[0044] In the formula, The lower surface attitude skew angle is... For arcsine conversion, For the half-size, The approximate response time constant is... The near-steady-state amplitude of the response. The preset combined RC calibration constant is used. As a preset power-law sensitivity factor, This refers to the minimum gap height.

[0045] In a preferred embodiment, the step of performing contact posture pre-compensation on the placement head based on the lower surface attitude skew angle and the minimum gap height to obtain the placement stroke parameters of the placement head includes:

[0046] Based on the lower surface attitude skew angle and the minimum gap height, the mounting head is pre-deflected in space attitude and pre-lifted in termination depth to obtain the pre-compensated mounting parameters of the mounting head.

[0047] Using the pre-compensated mounting parameters as the driving command, the lifting drive shaft of the mounting head is driven to descend. During the descent of the mounting head, the current pressure feedback value of the mounting head is captured in real time.

[0048] The current pressure feedback value is separated from the preset target pressure value, and the deviation obtained is converted into the driving voltage adjustment increment of the mounting head.

[0049] The incremental adjustment of the driving voltage is periodically fed back and superimposed on the driving voltage command of the mounting head until the current pressure feedback value reaches the target pressure value.

[0050] To address the aforementioned problems, this invention also provides a closed-loop control system for placement pressure in a pick-and-place machine based on multi-sensor fusion, the system comprising:

[0051] The signal acquisition module is used to acquire the vacuum pressure fundamental frequency and real-time speed of the placement head in the target placement machine, and to synchronously cut the uniform descent segment of the vacuum pressure fundamental frequency to obtain the vacuum pressure subsequence and vacuum pressure baseline value of the placement head.

[0052] The normalized pressure drop analysis module is used to perform normalized pressure drop analysis on the vacuum pressure subsequence with the vacuum pressure baseline value as a reference, so as to obtain the approximate pressure drop sequence of the mounting head.

[0053] The response feature extraction module is used to perform first-order inertial element time-domain feature fitting on the proximity pressure drop sequence to obtain the proximity response time constant and proximity response steady-state amplitude of the mounting head.

[0054] The gap height mapping module is used to map the steady-state amplitude of the proximity response to the gap height to obtain the minimum gap height between the component carried in the mounting head and the target substrate.

[0055] The skew angle inversion module is used to perform aerodynamic impedance inversion on the lower surface of the carrying element based on the proximity response time constant and the proximity response steady-state amplitude, so as to obtain the attitude skew angle of the lower surface of the carrying element.

[0056] The pre-compensation and pressure closed-loop module is used to perform pre-compensation of the contact posture of the mounting head based on the lower surface attitude skew angle and the minimum gap height, so as to obtain the mounting stroke parameters of the mounting head.

[0057] Compared with the prior art, the present invention has the following beneficial effects:

[0058] 1. This invention simultaneously acquires vacuum pressure time-series data and lifting shaft position time-series data, eliminates high-frequency ripples in the drive through notch filtering, and synchronously cuts the effective pressure waveform based on a uniform motion time window. Through normalized amplitude reduction analysis and time-domain fitting of a first-order inertial element, pressure response features are accurately extracted, enabling complete separation of effective pressure change information during the mounting approach phase and significantly improving the analytical accuracy of the vacuum pressure dynamic signal. Furthermore, by interpolating and mapping the calibration dataset to solve for the gap height, and combining this with an aerodynamic impedance model to quantitatively invert the component's lower surface attitude skew angle, the real-time contact state parameters between the component and the substrate can be accurately obtained.

[0059] 2. This invention performs pre-compensation of the contact posture of the placement head based on the gap height and attitude skew angle obtained by inversion, outputs placement stroke parameters adapted to the current component working conditions, and simultaneously collects pressure feedback in real time during the placement descent process. It periodically converts the pressure deviation into incremental closed-loop correction output commands for drive voltage adjustment, stably maintains the placement target pressure, effectively reduces process abnormalities caused by fluctuations in placement working conditions, and improves the process stability and product yield of the placement operation. Attached Figure Description

[0060] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the description of the embodiments of the present invention will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0061] Figure 1 This is a flowchart illustrating a closed-loop control method for placement pressure in a pick-and-place machine based on multi-sensor fusion, provided in an embodiment of the present invention.

[0062] Figure 2This is a functional block diagram of a pick-and-place machine mounting pressure closed-loop control system based on multi-sensor fusion provided in an embodiment of the present invention;

[0063] The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0064] It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

[0065] This application provides a closed-loop control method for placement pressure of a pick-and-place machine based on multi-sensor fusion. The executing entity of the closed-loop control method for placement pressure of a pick-and-place machine based on multi-sensor fusion includes, but is not limited to, at least one of the following electronic devices that can be configured to execute the method provided in this application: a server, a terminal, etc. In other words, the closed-loop control method for placement pressure of a pick-and-place machine based on multi-sensor fusion can be executed by software or hardware installed on a terminal device or a server device. The server includes, but is not limited to, a single server, a server cluster, a cloud server, or a cloud server cluster. The server can be an independent server or a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, content delivery networks (CDNs), and big data and artificial intelligence platforms.

[0066] Reference Figure 1 The diagram shown is a flowchart illustrating a closed-loop control method for placement pressure in a pick-and-place machine based on multi-sensor fusion, according to an embodiment of the present invention. In this embodiment, the closed-loop control method for placement pressure in a pick-and-place machine based on multi-sensor fusion includes:

[0067] S01. Obtain the vacuum pressure fundamental frequency and real-time speed of the placement head in the target placement machine, and synchronously cut the uniform descent segment of the vacuum pressure fundamental frequency to obtain the vacuum pressure subsequence and vacuum pressure baseline value of the placement head.

[0068] In this embodiment of the invention, obtaining the vacuum pressure fundamental frequency and real-time timing speed of the placement head in the target placement machine includes:

[0069] At the start of the current placement stroke, the pressure of the vacuum passage cavity of the placement head in the target pick and place machine is continuously sampled in the time domain to obtain the time-series vacuum pressure sequence of the placement head, and the position of the lifting drive shaft of the placement head is continuously sampled in the time domain to obtain the time-series position feedback sequence of the placement head.

[0070] Based on a preset notch frequency, the time-series vacuum pressure sequence is subjected to stopband filtering to eliminate the high-frequency drive ripple components superimposed in the time-series vacuum pressure sequence, while retaining the low-frequency trend components in the time-series vacuum pressure sequence, thereby obtaining the vacuum pressure fundamental frequency of the mounting head.

[0071] The displacement difference between adjacent sampling points in the time-series position feedback sequence is extracted, and the ratio between the displacement difference and the sampling time interval between the adjacent sampling points is used as the real-time speed of the mounting head.

[0072] The process of synchronously cutting the uniformly decreasing segment of the vacuum pressure fundamental frequency to obtain the vacuum pressure sub-sequence and vacuum pressure baseline value of the mounting head includes:

[0073] The starting and ending sampling points are calibrated to maintain a constant speed in the real-time timing sequence.

[0074] Based on the start sampling point time and the end sampling point time, all position sampling points are extracted from the time-series position feedback sequence, and the time span of all position sampling points is used as the uniform motion time window of the mounting head.

[0075] Based on the uniform motion time window, the continuous pressure waveform of the vacuum pressure fundamental frequency is cut out along the time axis to obtain the vacuum pressure sub-sequence of the mounting head;

[0076] Select a stable waveform in the vacuum pressure subsequence that has not yet shown a pressure drop trend, accumulate the pressure amplitude of the stable waveform, and use the quotient of the accumulated value and the total number of pressure amplitudes as the vacuum pressure baseline value of the mounting head.

[0077] At the time node corresponding to the start of the current placement stroke, pressure values ​​are continuously collected in the vacuum circuit cavity of the placement head inside the target placement machine. Each collection operation corresponds to a set of time markers and pressure values. After all collections are completed, they are integrated to form a time-series vacuum pressure sequence. The time-series vacuum pressure sequence refers to all pressure values ​​obtained in this complete collection operation and their respective matching time marker sets.

[0078] During the same time period when pressure acquisition is carried out simultaneously, position values ​​are continuously collected from the lifting drive shaft of the placement head. Each acquisition operation corresponds to a set of time markers and position values. After all acquisitions are completed, they are integrated to form a time-series position feedback sequence. The time-series position feedback sequence refers to all position values ​​obtained from this complete acquisition operation and their respective matching time marker sets.

[0079] The preset notch frequency is determined by the fixed fluctuation frequency generated by the normal operation of the pick-and-place machine's lifting drive shaft. The complete timing vacuum pressure sequence is sent to the band-stop signal filtering channel. The band-stop signal filtering channel will remove all fluctuation values ​​that match the preset notch frequency and only retain the pressure values ​​that do not match the preset notch frequency. All the pressure values ​​retained after filtering are combined with the corresponding time markers to form the vacuum pressure fundamental frequency. The vacuum pressure fundamental frequency refers to the complete pressure timing set that retains only the slow changing trend after removing the fluctuation values ​​caused by the drive.

[0080] The position values ​​of two adjacent sets within the time-series position feedback sequence are read sequentially. The displacement difference between the two sets of sampling points is obtained by subtracting the position value of the previous set from the position value of the latter set. The sampling time interval between adjacent sampling points is determined by the fixed acquisition action interval of the acquisition device. Each set of displacement difference is divided by the sampling time interval between adjacent sampling points in the same set. The results of each division operation are arranged and combined in the order of acquisition to form the time-series real-time speed. The time-series real-time speed refers to the set of values ​​of the speed of the lifting axis movement corresponding to each set of adjacent acquisition intervals.

[0081] Each set of velocity values ​​within the real-time velocity sequence is read sequentially. The specific magnitudes of multiple consecutive sets of velocity values ​​are compared. The first acquisition time when multiple consecutive sets of values ​​remain uniform is marked as the starting sampling point time. The acquisition time when the uniform and unchanging value ends is marked as the ending sampling point time. Both the starting sampling point time and the ending sampling point time are taken from the time stamps built into the real-time velocity sequence.

[0082] Read all position sampling points within the timing feedback sequence whose time markers are between the start and end sampling points. Define the complete duration from the start to the end time marker corresponding to all selected position sampling points as the uniform motion time window. The uniform motion time window refers to the complete acquisition duration interval corresponding to the lifting axis maintaining a constant moving speed.

[0083] Based on the start and end time markers corresponding to the uniform motion time window, select all continuous pressure values ​​within the time interval of the vacuum pressure fundamental frequency that fall within that time interval. Combine all the selected continuous pressure values ​​with the corresponding time markers to form a vacuum pressure subsequence. The vacuum pressure subsequence refers to the set of segmented pressure time sequences corresponding to the uniform motion phase of the lifting shaft.

[0084] The pressure values ​​within the vacuum pressure subsequence are read segment by segment according to time sequence. The continuous waveform segments where the values ​​do not show a continuous decreasing trend are distinguished. All pressure amplitudes within the segment are summed to obtain the total amplitude. The total number of pressure amplitudes contained in the stable waveform segment is counted. The sum of amplitudes is divided by the total number of pressure amplitudes to obtain a fixed value. This fixed value is the vacuum pressure baseline value. The vacuum pressure baseline value refers to the standard pressure value that remains stable and unchanged when the component is not close to the substrate.

[0085] The beneficial effects include the simultaneous acquisition of vacuum circuit pressure and lifting drive shaft position, resulting in two complete time series sets with corresponding time stamps. This enables a complete correspondence between pressure changes and shaft movement. Pressure signal filtering is performed using a notch filter frequency determined by the inherent fluctuations of the lifting drive shaft, removing interference values ​​caused by shaft movement and retaining only the slow pressure change trend. This ensures that the pressure-related acquisition content only retains the effective change information caused by the component's approach. Based on the position time series values, the shaft movement speed time series sets are derived, accurately locating the complete time interval corresponding to uniform shaft movement. Segmented pressure time series sets are extracted within this time period, and stable pressure values ​​that do not change are selected to obtain standard pressure values. This completely separates the stable pressure benchmark during the stage before the component approaches the substrate, allowing subsequent pressure analysis for component approach to rely solely on interference-free, clearly defined segmented pressures, ensuring that subsequent pressure-related analysis has complete and reliable raw data support.

[0086] S02. Using the vacuum pressure baseline value as a reference, perform normalized pressure drop analysis on the vacuum pressure subsequence to obtain the approximate pressure drop sequence of the mounting head.

[0087] In this embodiment of the invention, the step of performing normalized pressure drop analysis on the vacuum pressure subsequence using the vacuum pressure baseline value as a reference to obtain the approximate pressure drop sequence of the mounting head includes:

[0088] The net pressure change of the vacuum pressure subsequence is obtained by subtracting the amplitude of the sampling points in the vacuum pressure subsequence from the baseline value of the vacuum pressure bit by bit.

[0089] Using the vacuum pressure baseline value as a reference, the net pressure change is divided by the vacuum pressure baseline value, and the quotients are arranged sequentially to obtain the proximity pressure drop sequence of the mounting head.

[0090] The amplitude values ​​of each sampling point within the vacuum pressure subsequence, arranged sequentially according to time markers, are extracted in turn. The vacuum pressure subsequence refers to the set of segmented pressure time sequences corresponding to the uniform movement stage of the lifting shaft. The sampling point amplitude refers to the pressure value matched by each set of time markers within the vacuum pressure subsequence. The vacuum pressure baseline value is subtracted from the amplitude value of each sampling point. The vacuum pressure baseline value refers to the standard pressure value that remains stable and unchanged when the component is not close to the substrate. Each subtraction operation yields the numerical result at the corresponding position. All numerical results are completely retained in the original time arrangement order to form the net pressure change. The net pressure change refers to the difference between each pressure value within the vacuum pressure subsequence and the standard pressure value.

[0091] Each set of differences within the net pressure change is extracted sequentially according to time. Each set of differences is divided by the vacuum pressure baseline value, which refers to the standard pressure value that remains stable before the component approaches the substrate. Each division operation yields the ratio at the corresponding position. All ratios are arranged and integrated in strict accordance with the original time sequence of the net pressure change. After integration, a sequence of pressure drop approaches is formed. The sequence of pressure drop approaches refers to the complete time sequence content composed of all ratios and corresponding time markers.

[0092] The beneficial effects are that, based on the baseline value of vacuum pressure, the difference in amplitude of all sampling points within the vacuum pressure subsequence is used to obtain the net pressure change, completely removing the fixed numerical component brought by the reference pressure, and only retaining the pressure change caused by the component's approach to the substrate. Then, the ratio conversion is completed using the same reference to generate a near pressure drop sequence, unifying the numerical measurement scale under different initial reference pressures, eliminating the numerical comparison obstacles caused by the difference in initial reference pressures, and completely preserving the temporal arrangement order to maintain the complete trend of pressure change over time. This provides original material with a unified scale and complete temporal sequence for subsequent extraction of pressure change-related content, so that subsequent analysis of pressure change during the component's approach process is not affected by the initial vacuum reference value.

[0093] S03. Perform first-order inertial element time-domain feature fitting on the approach pressure drop sequence to obtain the approach response time constant and approach response steady-state amplitude of the mounting head.

[0094] In this embodiment of the invention, the step of performing first-order inertial element time-domain feature fitting on the proximity pressure drop sequence to obtain the proximity response time constant and proximity response steady-state amplitude of the mounting head includes:

[0095] By backtracking point by point from the starting point of the approaching pressure drop sequence, the tail convergence segment is obtained;

[0096] The pressure drop values ​​of all sampling points within the tail convergence section are accumulated and summarized, and the quotient between the accumulated summation result and the total number of sampling points within the tail convergence section is taken as the proximity response steady-state amplitude of the mounting head.

[0097] Starting from the beginning of the approach pressure drop sequence, the difference between the pressure drop value at the sampling point in the approach pressure drop sequence and the pressure drop value at the beginning point is statistically analyzed point by point along the positive direction of the time axis.

[0098] When the difference first reaches a specified percentage of the near-response steady-state amplitude, the characteristic sampling position is obtained;

[0099] The time span between the feature sampling position and the starting point is used as the proximity response time constant of the mounting head.

[0100] The step of backtracking point by point from the starting point of the near-pressure drop sequence to obtain the tail convergence segment includes:

[0101] Starting from the beginning of the approach pressure drop sequence, traverse backward point by point to extract the variation between each sampling point in the approach pressure drop sequence and the previous sampling point.

[0102] Define a continuous segment where the change amount is first continuously lower than a preset tolerance limit, and mark the starting boundary of the continuous segment as the convergence start point;

[0103] All sampling points between the convergence start point and the termination point of the near pressure drop sequence are defined as the tail convergence segment.

[0104] The sampling point closest to the end of the corresponding time series within the pressure drop sequence is taken as the tail start point. The "closest pressure drop sequence" refers to the complete time series content composed of all ratios and corresponding time markers. From the tail start point towards the beginning of the time series, each sampling point is retrieved sequentially to complete the reverse point-by-point traversal operation.

[0105] The pressure drop values ​​corresponding to each of the two adjacent sampling points are retrieved during the traversal process. The pressure drop value of the retrieved sampling point is then subtracted from the pressure drop value of the previous sampling point to obtain the change between the two sampling points. The change refers to the difference in pressure drop values ​​between adjacent sampling points.

[0106] The preset tolerance limit is determined based on the allowable range of pressure value fluctuations during the stable stage of the mounting operation. Each set of changes is continuously compared with the preset tolerance limit. A continuous segment in which all sets of changes meet the corresponding value standard is recorded. The boundary position of this continuous segment towards the earlier side of the time sequence is marked as the convergence start point. The convergence start point refers to the time sequence boundary position where the pressure value no longer changes continuously.

[0107] Read all sampling points whose time markers are within the convergence start point to the end point of the time sequence, and integrate all the selected sampling points to define the tail convergence segment. The tail convergence segment refers to the time sequence segment where the pressure value remains stable after the component approaches the action.

[0108] The pressure drop values ​​corresponding to all sampling points within the tail convergence section are retrieved. All pressure drop values ​​are summed up sequentially to obtain a uniform total value. The total number of sampling points within the tail convergence section is counted, and the total value is distributed to each sampling point to obtain a uniform value. This uniform value is the steady-state amplitude of the proximity response. The steady-state amplitude of the proximity response refers to the standard value of the pressure drop that remains stable after the element completes the proximity action.

[0109] The sampling point closest to the beginning of the pressure drop sequence is taken as the starting point. The starting point has its own pressure drop value. The pressure drop value corresponding to each sampling point is retrieved step by step along the time sequence. The pressure drop value of the starting point is subtracted from the pressure drop value of each sampling point to obtain the corresponding position difference.

[0110] The preset percentage is determined based on the time domain determination rules of the first-order inertial system. The preset percentage is fixed at 63.2. The corresponding value obtained after continuously comparing the position difference with the near-response steady-state amplitude of 63.2 is recorded. The sampling point position that first meets the corresponding value standard is recorded as the feature sampling position. The feature sampling position refers to the time boundary point where the pressure change reaches the stable standard.

[0111] Extract the time stamps inherent to the feature sampling location and the time stamps inherent to the starting point, and compare the complete time interval covered by the two sets of time stamps. This time interval is the proximity response time constant, which refers to the time consumed from the initial approach action of the element to the pressure value stabilizing.

[0112] The beneficial effects are as follows: by traversing and comparing the changes in adjacent sampling points in reverse, a stable tail convergence segment is defined; only segments where the values ​​no longer fluctuate are used to obtain the steady-state amplitude of the approximate response, avoiding interference from dynamically changing data on the standard values; all stable information within the segment is integrated by summarizing and evenly distributing all sampled values ​​within the segment, ensuring that the steady-state amplitude of the approximate response matches the actual pressure drop state after the component approaches; the characteristic sampling position is located based on a fixed judgment standard; the approximate response time constant is determined based on the time series mark difference; both characteristic values ​​are obtained by layer-by-layer filtering based on complete time series segments; and fluctuating data is filtered layer by layer based on the original time series content throughout the process, so that the two types of response characteristic values ​​have a unified and reliable data source, providing an accurate and stable basic reference for subsequent analysis of component gaps and attitudes.

[0113] S04. Map the proximity response steady-state amplitude to obtain the minimum gap height between the component carried in the mounting head and the target substrate.

[0114] In this embodiment of the invention, the step of mapping the proximity response steady-state amplitude to obtain the minimum gap height between the component carried in the mounting head and the target substrate includes:

[0115] Based on multiple sets of known gap heights, the vacuum pressure response of the target placement machine is measured to obtain the measured steady-state amplitude. The multiple sets of known gap heights and the measured steady-state amplitude are then compiled into a mapping record set.

[0116] Using the near-response steady-state amplitude as a matching benchmark, the mapping record set is traversed, and the absolute value of the difference between the near-response steady-state amplitude and the measured steady-state amplitude is calculated.

[0117] The minimum value is selected from the absolute values ​​of the differences, and the known gap height of the minimum value is used as the reference height.

[0118] Using the reference height as the positioning center, extract the upper neighbor height and lower neighbor height of the reference height from the mapping record set;

[0119] Based on the steady-state amplitude of the proximity response, the upper neighbor height and the lower neighbor height are linearly proportionally interpolated to obtain the minimum gap height.

[0120] Multiple sets of fixed gap height values ​​are sequentially set (micron-level gap setting is achieved using laser displacement sensor calibration; the test conditions are consistent with the actual mounting conditions, including the same component model, reference vacuum level, and descent speed). Each set of gap height values ​​corresponds to a complete vacuum pressure response measurement operation. Each measurement operation produces a set of measured steady-state amplitude values. The measured steady-state amplitude values ​​refer to the standard value of stable pressure drop obtained under fixed gap height conditions. Each set of gap height values ​​is bound one-to-one with the corresponding measured steady-state amplitude values. All bound pairs are uniformly collected and integrated to form a mapping record set. The mapping record set refers to the complete set of content where gap height and matching steady-state amplitude are stored in pairs.

[0121] The steady-state amplitude of the proximity response is retrieved as a unified comparison reference. The steady-state amplitude of the proximity response refers to the standard value of the pressure drop that remains stable after the component completes the proximity action. The measured steady-state amplitudes stored in pairs within the mapping record set are retrieved in sequence. The difference between the steady-state amplitude of the proximity response and each set of measured steady-state amplitudes is calculated. The negative values ​​produced by each difference calculation operation are converted into positive values ​​and stored. All positive values ​​are stored to form the absolute value of the difference. The absolute value of the difference refers to the difference that is not distinguishable between positive and negative values ​​generated by comparing the two sets of steady-state amplitudes.

[0122] Read all the retained absolute difference values ​​completely, compare the values ​​corresponding to each group of absolute difference values ​​one by one, filter out the absolute difference values ​​that are uniquely at the lowest level after comparing the values ​​with all other absolute difference values, retrieve the known gap height that is bound to the group of absolute difference values, and the known gap height that is bound to the group of absolute difference values ​​is the reference height. The reference height refers to the record matching gap height that has the highest degree of matching with the current near-response steady-state amplitude.

[0123] Retrieve all paired records within the mapping record set, locate the record entry corresponding to the reference height, read the gap height of the preceding entry in the record arrangement order as the upper neighbor height, and read the gap height of the following entry in the record arrangement order as the lower neighbor height. The upper neighbor height and the lower neighbor height refer to the two sets of gap height records arranged adjacent to the reference height.

[0124] Using the near-response steady-state amplitude as the allocation reference, two numerical intervals are divided by comparing the measured steady-state amplitudes of the upper and lower adjacent heights. The components corresponding to the two gap heights are allocated according to the position of the near-response steady-state amplitude within the interval. The components of the two gap heights are integrated to obtain a single fixed value, which is the minimum gap height. The minimum gap height refers to the actual gap value between the current component and the substrate.

[0125] The beneficial effects are that multiple pairs of gap heights are collected in advance and integrated with the measured steady-state amplitudes into a mapping record set, establishing a correspondence between pressure stability values ​​and gap states. The real-time acquired near-response steady-state amplitudes are used to traverse all records to complete the difference comparison, accurately locating the reference height with the highest matching degree. Based on the two sets of gap heights adjacent to the reference height, the interval is divided to complete the component integration to obtain the minimum gap height. It is no longer limited to the fixed gap values ​​that have been pre-entered in the record set, and can output gaps that fit the real-time working conditions. The gap solution is completed entirely based on the correspondence left by the actual measurement, so that the acquisition of gap values ​​is supported by on-site measured data, ensuring that the final minimum gap height fits the actual spatial state between the component and the substrate.

[0126] S05. Based on the proximity response time constant and the proximity response steady-state amplitude, perform aerodynamic impedance inversion on the lower surface of the carrying element to obtain the attitude deflection angle of the lower surface of the carrying element.

[0127] In this embodiment of the invention, the step of performing aerodynamic impedance inversion on the lower surface of the carrying element based on the proximity response time constant and the steady-state amplitude of the proximity response to obtain the attitude skew angle of the lower surface of the carrying element includes:

[0128] Obtain the outline dimensions of the component carried in the mounting head, and perform a half-section of the outline dimensions along the symmetrical midline of the component to obtain the half-size of the component.

[0129] Based on the proximity response time constant and the half-size, the lower surface attitude skew angle of the carrying element is calculated, wherein the formula for calculating the lower surface attitude skew angle is:

[0130] ;

[0131] In the formula, The lower surface attitude skew angle is... For arcsine conversion, For the half-size, The approximate response time constant is... The near-steady-state amplitude of the response. The preset combined RC calibration constant is used. As a preset power-law sensitivity factor, This refers to the minimum gap height.

[0132] The entire outer contour of the component currently adsorbed and fixed by the mounting head is collected, and the outer contour dimensions refer to the set of length values ​​corresponding to all external boundaries of the component. The symmetrical mid-plane of the component's own structure is located. The symmetrical mid-plane refers to the dividing plane on both sides of the component's outer contour structure that are completely corresponding. The entire outer contour length corresponding to one side is cut off with this dividing plane as the dividing boundary. After the cutting is completed, it is integrated to form a half-size. The half-size refers to the set of outer contour lengths corresponding to one side of the component after being divided along the symmetrical mid-plane.

[0133] The system retrieves two types of data from the previous process: the proximity response time constant and the half-size. The proximity response time constant refers to the time it takes for the element to complete its approach motion and for the pressure value to stabilize. The half-size refers to the set of lengths of the element on one side after it is divided along the symmetrical bisecting plane. Both types of data are simultaneously fed into the aerodynamic impedance inversion execution process. Aerodynamic impedance inversion refers to the complete processing process that derives the tilt state of the element's bottom surface based on the time-series changes in vacuum pressure and the element's dimensions. It also incorporates the steady-state amplitude of the proximity response from the previous process into the complete processing process. Throughout the process, it matches the length of the element's one side with the aerodynamic conduction state corresponding to the time-series changes in vacuum pressure. It derives the degree of tilt of the element's lower surface relative to the standard placement state layer by layer, and finally produces a unique fixed value, which is the lower surface attitude skew angle. The lower surface attitude skew angle refers to the degree of tilt of the lower surface of the element that deviates from the horizontal standard state.

[0134] The term "half-size" refers to the complete acquisition of the external dimensions corresponding to the entire outer contour of the component currently adsorbed and fixed by the mounting head. The symmetrical dividing plane of the component's own structure is used as the dividing boundary to cut off the entire length of the corresponding shape on one side. After the cutting is completed, the components are integrated to form the half-size.

[0135] The approach response time constant is defined by taking the earliest sampling point in the approach pressure drop sequence as the starting point and then progressively taking the pressure drop value corresponding to each sampling point along the time sequence. The pressure drop value at each sampling point is subtracted from the pressure drop value at the starting point to obtain the corresponding position difference. The preset percentage is determined according to the time domain judgment rule of the first-order inertial system, and this preset percentage is fixed at 63.2%. The corresponding values ​​obtained after matching each position difference with the steady-state amplitude of the approach response to 63.2 are continuously compared. The sampling point position that first meets the corresponding value standard is recorded as the feature sampling position. The time stamp of the feature sampling position is extracted and compared with the time stamp of the starting point. The complete time interval covered by the two sets of time stamps is compared, and this time interval is the approach response time constant.

[0136] This refers to the near-steady-state amplitude of the response. The sampling point at the end of the corresponding time sequence within the near-pressure drop sequence is taken as the tail start point. Starting from the tail start point and moving towards the beginning of the time sequence, each sampling point is retrieved sequentially to complete a reverse point-by-point traversal. During the traversal, the pressure drop values ​​corresponding to each of two adjacent sets of sampling points are retrieved. The pressure drop value of the retrieved sampling point is subtracted from the pressure drop value of the previous set of sampling points to obtain the variation between the two sets of sampling points. The preset tolerance limit is determined based on the allowable pressure value fluctuation range during the stable phase of the mounting operation. Each set of variations is continuously compared with the preset tolerance limit, and records are kept until multiple consecutive sets of variations meet the requirement. For a continuous segment of the numerical standard, the boundary position of this continuous segment towards the earlier side of the time series is marked as the convergence start point. All sampling points within the pressure drop sequence whose time markers are between the convergence start point and the end point of the time series are read. All the selected sampling points are integrated and defined as the tail convergence segment. The pressure drop values ​​corresponding to each sampling point within the tail convergence segment are retrieved. All pressure drop values ​​are summed up sequentially to obtain a unified numerical sum. The total number of sampling points contained within the tail convergence segment is counted. The numerical sum is then distributed to each sampling point to obtain a unified value. This unified value is the near-response steady-state amplitude.

[0137] The reference value is the comprehensive calibration constant of resistance and capacitance. It is determined by completing multiple calibration tests based on the length of the vacuum air path of the chip mounter, the inner diameter of the pipe, and the output specifications of the vacuum generator. The unified and fixed value produced by the multiple tests is used as the comprehensive calibration constant of resistance and capacitance.

[0138] The power-law sensitive factor is uniformly calibrated based on the measured data of vacuum pressure response corresponding to multiple sets of elements with different tilt degrees. The fixed value produced after calibration is used as the power-law sensitive factor.

[0139] The minimum gap height is defined by pre-collecting multiple sets of gap heights and matching measured steady-state amplitudes to construct a mapping record set. The difference between the real-time acquired near-response steady-state amplitudes and the records set is compared. The reference height with the highest matching degree is selected, and the minimum gap height is obtained by extracting the gap heights of two adjacent sets of reference heights, dividing the intervals, allocating components, and integrating them.

[0140] The corresponding arcsine conversion reads the unified value produced by the integration of all previous processes and sends the unified value into the arcsine conversion process. The arcsine conversion process retrieves the standard angle correspondence table and matches the unique angle corresponding to the input value. This angle is the lower surface attitude skew angle.

[0141] The entire numerical processing workflow synchronously integrates the inherent dimensions of the mounted components, two types of response characteristic values ​​generated by the temporal changes in vacuum pressure, the real-time spatial gap between the components and the substrate, and the fixed values ​​of vacuum conduction calibration in the gas path. Based on the temporal changes in vacuum pressure as the components approach, the workflow matches the single-sided dimensions of the components to complete the derivation of the tilt state of the component's bottom surface. The entire processing workflow does not add any additional optical or contact detection components. It only relies on the vacuum pressure temporal data collected in situ during the mounting operation and the inherent dimensions of the components to accurately determine the tilt degree of the component's bottom surface. The unique angle produced completely corresponds to the tilt degree of the component's lower surface deviating from the horizontal standard placement state, providing an accurate tilt state reference for subsequent pre-compensation operations of the mounting head's contact posture.

[0142] When the length of the single-sided outline of the component corresponding to the half-size increases, the angle produced by the final output of the entire numerical processing flow continues to narrow. When the time it takes for the component to approach the pressure stabilization point, which is close to the response time constant, increases, the intermediate value produced by the pre-integration within the entire numerical processing flow continues to expand. When the standard value of the stable pressure drop, which is close to the steady-state amplitude of the response, expands, the intermediate value produced by the pre-integration within the entire numerical processing flow continues to expand. When the fixed value of the gas path calibration, which is close to the resistance-capacitance comprehensive calibration constant, expands, the intermediate value produced by the pre-integration within the entire numerical processing flow continues to expand. When the fixed value of the calibration, which is close to the power-law sensitive factor, changes, the intermediate value and the final angle within the entire numerical processing flow synchronously produce corresponding matching changes. When the component substrate gap value, which is close to the minimum gap height, expands, the intermediate value produced by the pre-integration within the entire numerical processing flow continues to narrow. The arcsine conversion process matches the input intermediate value to complete the angle conversion, and the direction of change of the input intermediate value is completely consistent with the direction of change of the final output angle.

[0143] The beneficial effects are: complete acquisition of all component dimensions and extraction of half-size by cutting off one side along the symmetrical midline; standardized extraction of dimensions based on the component's own geometric boundaries; elimination of redundant information from the complete external dimensions; matching the data acquisition requirements of pneumatic conduction related processing flows; simultaneous pneumatic impedance inversion by combining the two types of pressure response related content from previous processes; deduction of the component's bottom surface tilt degree based on the vacuum pressure time series change and the component's one-side external dimensions; attitude determination based on the real-time aerodynamic time series data and the component's inherent shape information acquired during the entire mounting process; and the lower surface attitude deviation angle can be obtained without the need for additional external detection devices, providing a reference for the actual placement state of the component for subsequent pre-positioning compensation of the mounting head attitude.

[0144] S06. Based on the lower surface attitude skew angle and the minimum gap height, perform contact attitude pre-compensation on the mounting head to obtain the mounting stroke parameters of the mounting head.

[0145] In this embodiment of the invention, the step of performing contact posture pre-compensation on the placement head based on the lower surface attitude skew angle and the minimum gap height to obtain the placement stroke parameters of the placement head includes:

[0146] Based on the lower surface attitude skew angle and the minimum gap height, the mounting head is pre-deflected in space attitude and pre-lifted in termination depth to obtain the pre-compensated mounting parameters of the mounting head.

[0147] Using the pre-compensated mounting parameters as the driving command, the lifting drive shaft of the mounting head is driven to descend. During the descent of the mounting head, the current pressure feedback value of the mounting head is captured in real time.

[0148] The current pressure feedback value is separated from the preset target pressure value, and the deviation obtained is converted into the driving voltage adjustment increment of the mounting head.

[0149] The incremental adjustment of the driving voltage is periodically fed back and superimposed on the driving voltage command of the mounting head until the current pressure feedback value reaches the target pressure value.

[0150] The system retrieves two types of data from the previous process: the lower surface attitude skew angle and the minimum gap height. The lower surface attitude skew angle refers to the degree of tilt of the lower surface of the component that deviates from the horizontal standard state. The minimum gap height refers to the actual gap value between the component and the substrate. The placement angle of the mounting head is adjusted according to the tilt degree corresponding to the lower surface attitude skew angle to complete the spatial attitude pre-deflection operation. The height position of the mounting head at the end of the descent is adjusted according to the spatial distance corresponding to the minimum gap height to complete the termination depth pre-lift operation. After the two adjustment operations are completed, the pre-compensation mounting parameters are generated. The pre-compensation mounting parameters refer to the mounting head angle and the descent end height that are adapted to the tilt and gap state of the component.

[0151] The pre-compensation placement parameters are retrieved as the basis for the action execution of the lifting drive shaft. The pre-compensation placement parameters refer to the placement head angle and the height of the descent endpoint, which are related to the tilt and gap state of the adapter components. According to the angle and height information recorded in the pre-compensation placement parameters, the lifting drive shaft is controlled to drive the placement head to move downward at a constant speed to complete the placement descent drive operation. The pressure value inside the vacuum circuit is collected synchronously throughout the continuous downward movement of the placement head. The pressure value collected each time is called the current pressure feedback value. The current pressure feedback value refers to the real-time air circuit pressure collected during the placement descent process.

[0152] The preset target pressure value is determined based on the bonding pressure standard specified in the standard patch process. The current pressure feedback value and the preset target pressure value are retrieved. The deviation removal operation is completed by subtracting the preset target pressure value from the current pressure feedback value. The difference produced by the deviation removal operation is uniformly called the deviation amount. The deviation amount refers to the difference between the real-time collected pressure and the process standard pressure. The value corresponding to the deviation amount is matched with the lifting drive shaft voltage adjustment standard, and the deviation amount is completely converted into the drive voltage adjustment increment. The drive voltage adjustment increment refers to the voltage change used to correct the lifting drive shaft's movement force.

[0153] The drive voltage command that the placement head originally used to perform the descent action is retrieved. The drive voltage adjustment increment refers to the voltage variation used to correct the force of the lifting drive shaft. The drive voltage adjustment increment is superimposed on the original drive voltage command according to a fixed acquisition repetition cycle. After superposition, a new drive voltage command is generated to control the lifting drive shaft to continuously drive the placement head to move downward. The current pressure feedback value is collected synchronously throughout the process, and the deviation stripping and voltage superposition operations are repeated until the current pressure feedback value collected is completely consistent with the preset target pressure value. The entire set of angle and height control contents that are stably matched to the target pressure after the entire process of superposition and correction is integrated to form the placement stroke parameters. The placement stroke parameters refer to the complete set of control contents corresponding to the complete descent action of the placement head after the attitude compensation and pressure correction are completed.

[0154] The beneficial effects are that, based on the component tilt and gap values ​​produced by the preprocess, the placement head angle and descent endpoint height are simultaneously pre-adjusted, generating pre-compensated placement parameters adapted to real-time working conditions. This eliminates the bonding position deviation caused by component tilt and spatial gap in advance. The pre-compensation drives the placement head to descend while simultaneously collecting air pressure throughout the process. The difference is extracted and converted into voltage variation content by comparing it with the standard pressure specified in the process. The variation content is periodically superimposed to correct the drive command and continuously adjust the placement action until the real-time pressure matches the process standard. The entire set of control content is integrated into placement stroke parameters, and the attitude pre-correction and pressure dynamic correction are completed at the same time. This avoids the problem of bonding pressure deviation from the standard caused by fixed stroke and stably maintains the standard pressure state of the bonding operation.

[0155] like Figure 2 The diagram shown is a functional block diagram of a chip mounter mounting pressure closed-loop control system based on multi-sensor fusion provided in an embodiment of the present invention.

[0156] The multi-sensor fusion-based pick-and-place machine pressure closed-loop control system 100 described in this invention can be installed in an electronic device. Depending on the functions implemented, the multi-sensor fusion-based pick-and-place machine pressure closed-loop control system 100 may include a signal acquisition module 101, a normalized amplitude reduction analysis module 102, a response feature extraction module 103, a gap height mapping module 104, a skew angle inversion module 105, and a pre-compensation and pressure closed-loop module 106. The module described in this invention can also be called a unit, referring to a series of computer program segments that can be executed by the processor of an electronic device and perform a fixed function, stored in the memory of the electronic device.

[0157] In this embodiment, the functions of each module / unit are as follows:

[0158] The signal acquisition module 101 is used to acquire the vacuum pressure fundamental frequency and real-time speed of the mounting head in the target pick and place machine, and to synchronously cut the uniform descent segment of the vacuum pressure fundamental frequency to obtain the vacuum pressure subsequence and vacuum pressure baseline value of the mounting head.

[0159] The normalized pressure drop analysis module 102 is used to perform normalized pressure drop analysis on the vacuum pressure subsequence with the vacuum pressure baseline value as a reference, so as to obtain the approximate pressure drop sequence of the mounting head.

[0160] The response feature extraction module 103 is used to perform first-order inertial element time-domain feature fitting on the approach pressure drop sequence to obtain the approach response time constant and approach response steady-state amplitude of the mounting head.

[0161] The gap height mapping module 104 is used to map the proximity response steady-state amplitude to the gap height to obtain the minimum gap height between the component carried in the mounting head and the target substrate.

[0162] The skew angle inversion module 105 is used to perform aerodynamic impedance inversion on the lower surface of the carrying element based on the proximity response time constant and the proximity response steady-state amplitude, so as to obtain the attitude skew angle of the lower surface of the carrying element.

[0163] The pre-compensation and pressure closed-loop module 106 is used to perform pre-compensation of the contact posture of the mounting head based on the lower surface attitude skew angle and the minimum gap height, so as to obtain the mounting stroke parameters of the mounting head.

[0164] In the several embodiments provided by this invention, it should be understood that the disclosed methods and systems can be implemented in other ways. For example, the system embodiments described above are merely illustrative; for instance, the division of modules is only a logical functional division, and other division methods may be used in actual implementation.

[0165] Furthermore, the functional modules in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or in the form of hardware plus software functional modules.

[0166] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0167] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A closed-loop control method for placement pressure in a pick-and-place machine based on multi-sensor fusion, characterized in that, The method includes: S01. Obtain the vacuum pressure fundamental frequency and real-time speed of the placement head in the target placement machine, and synchronously cut the uniform descent segment of the vacuum pressure fundamental frequency to obtain the vacuum pressure subsequence and vacuum pressure baseline value of the placement head. S02. Using the vacuum pressure baseline value as a reference, perform normalized pressure drop analysis on the vacuum pressure subsequence to obtain the approximate pressure drop sequence of the mounting head. S03. Perform first-order inertial element time-domain feature fitting on the approach pressure drop sequence to obtain the approach response time constant and approach response steady-state amplitude of the mounting head. S04. Map the proximity response steady-state amplitude to obtain the minimum gap height between the component carried in the mounting head and the target substrate. S05. Based on the proximity response time constant and the proximity response steady-state amplitude, perform aerodynamic impedance inversion on the lower surface of the carrying element to obtain the attitude deflection angle of the lower surface of the carrying element. S06. Based on the lower surface attitude skew angle and the minimum gap height, perform contact attitude pre-compensation on the mounting head to obtain the mounting stroke parameters of the mounting head.

2. The placement pressure closed-loop control method for a pick-and-place machine based on multi-sensor fusion as described in claim 1, characterized in that, The acquisition of the vacuum pressure fundamental frequency and real-time timing speed of the placement head in the target placement machine includes: At the start of the current placement stroke, the pressure of the vacuum passage cavity of the placement head in the target pick and place machine is continuously sampled in the time domain to obtain the time-series vacuum pressure sequence of the placement head, and the position of the lifting drive shaft of the placement head is continuously sampled in the time domain to obtain the time-series position feedback sequence of the placement head. Based on a preset notch frequency, the time-series vacuum pressure sequence is subjected to stopband filtering to eliminate the high-frequency drive ripple components superimposed in the time-series vacuum pressure sequence, while retaining the low-frequency trend components in the time-series vacuum pressure sequence, thereby obtaining the vacuum pressure fundamental frequency of the mounting head. The displacement difference between adjacent sampling points in the time-series position feedback sequence is extracted, and the ratio between the displacement difference and the sampling time interval between the adjacent sampling points is used as the real-time speed of the mounting head.

3. The placement pressure closed-loop control method for a pick-and-place machine based on multi-sensor fusion as described in claim 1, characterized in that, The process of synchronously cutting the uniformly decreasing segment of the vacuum pressure fundamental frequency to obtain the vacuum pressure sub-sequence and vacuum pressure baseline value of the mounting head includes: The starting and ending sampling points are calibrated to maintain a constant speed in the real-time timing sequence. Based on the start sampling point time and the end sampling point time, all position sampling points are extracted from the timing position feedback sequence of the mounting head, and the time span of all position sampling points is used as the uniform motion time window of the mounting head. Based on the uniform motion time window, the continuous pressure waveform of the vacuum pressure fundamental frequency is cut out along the time axis to obtain the vacuum pressure sub-sequence of the mounting head; Select a stable waveform in the vacuum pressure subsequence that has not yet shown a pressure drop trend, accumulate the pressure amplitude of the stable waveform, and use the quotient of the accumulated value and the total number of pressure amplitudes as the vacuum pressure baseline value of the mounting head.

4. The placement pressure closed-loop control method for a pick-and-place machine based on multi-sensor fusion as described in claim 1, characterized in that, The process involves using the vacuum pressure baseline as a reference to perform normalized pressure drop analysis on the vacuum pressure subsequence, resulting in the approximate pressure drop sequence of the mounting head, including: The net pressure change of the vacuum pressure subsequence is obtained by subtracting the amplitude of the sampling points in the vacuum pressure subsequence from the baseline value of the vacuum pressure bit by bit. Using the vacuum pressure baseline value as a reference, the net pressure change is divided by the vacuum pressure baseline value, and the quotients are arranged sequentially to obtain the proximity pressure drop sequence of the mounting head.

5. The placement pressure closed-loop control method for a pick-and-place machine based on multi-sensor fusion as described in claim 1, characterized in that, The step of fitting the proximity pressure drop sequence with the time-domain characteristics of a first-order inertial element to obtain the proximity response time constant and the proximity response steady-state amplitude of the mounting head includes: By backtracking point by point from the starting point of the approaching pressure drop sequence, the tail convergence segment is obtained; The pressure drop values ​​of all sampling points within the tail convergence section are accumulated and summarized, and the quotient between the accumulated summation result and the total number of sampling points within the tail convergence section is taken as the proximity response steady-state amplitude of the mounting head. Starting from the beginning of the approach pressure drop sequence, the difference between the pressure drop value at the sampling point in the approach pressure drop sequence and the pressure drop value at the beginning point is statistically analyzed point by point along the positive direction of the time axis. When the difference first reaches a specified percentage of the near-response steady-state amplitude, the characteristic sampling position is obtained; The time span between the feature sampling position and the starting point is used as the proximity response time constant of the mounting head.

6. The placement pressure closed-loop control method for a pick-and-place machine based on multi-sensor fusion as described in claim 5, characterized in that, The step of backtracking point by point from the starting point of the near-pressure drop sequence to obtain the tail convergence segment includes: Starting from the beginning of the approach pressure drop sequence, traverse backward point by point to extract the variation between each sampling point in the approach pressure drop sequence and the previous sampling point. Define a continuous segment where the change amount is first continuously lower than a preset tolerance limit, and mark the starting boundary of the continuous segment as the convergence start point; All sampling points between the convergence start point and the termination point of the near pressure drop sequence are defined as the tail convergence segment.

7. The placement pressure closed-loop control method for a pick-and-place machine based on multi-sensor fusion as described in claim 1, characterized in that, The step of mapping the proximity response steady-state amplitude to the minimum gap height between the component carried in the mounting head and the target substrate includes: Based on multiple sets of known gap heights, the vacuum pressure response of the target placement machine is measured to obtain the measured steady-state amplitude. The multiple sets of known gap heights and the measured steady-state amplitude are then compiled into a mapping record set. Using the near-response steady-state amplitude as a matching benchmark, the mapping record set is traversed, and the absolute value of the difference between the near-response steady-state amplitude and the measured steady-state amplitude is calculated. The minimum value is selected from the absolute values ​​of the differences, and the known gap height of the minimum value is used as the reference height. Using the reference height as the positioning center, extract the upper neighbor height and lower neighbor height of the reference height from the mapping record set; Based on the steady-state amplitude of the proximity response, the upper neighbor height and the lower neighbor height are linearly proportionally interpolated to obtain the minimum gap height.

8. The placement pressure closed-loop control method for a pick-and-place machine based on multi-sensor fusion as described in claim 1, characterized in that, The step of performing aerodynamic impedance inversion on the lower surface of the carrier element based on the proximity response time constant and the steady-state amplitude of the proximity response to obtain the attitude deflection angle of the lower surface of the carrier element includes: Obtain the outline dimensions of the component carried in the mounting head, and perform a half-section of the outline dimensions along the symmetrical midline of the component to obtain the half-size of the component. Based on the proximity response time constant and the half-size, the lower surface attitude skew angle of the carrying element is calculated, wherein the formula for calculating the lower surface attitude skew angle is: ； In the formula, The lower surface attitude skew angle is... For arcsine conversion, For the half-size, The approximate response time constant is... The near-steady-state amplitude of the response. The preset combined RC calibration constant is used. As a preset power-law sensitivity factor, This refers to the minimum gap height.

9. The placement pressure closed-loop control method for a pick-and-place machine based on multi-sensor fusion as described in claim 1, characterized in that, The method of performing contact posture pre-compensation on the mounting head based on the lower surface attitude skew angle and the minimum gap height to obtain the mounting stroke parameters of the mounting head includes: Based on the lower surface attitude skew angle and the minimum gap height, the mounting head is pre-deflected in space attitude and pre-lifted in termination depth to obtain the pre-compensated mounting parameters of the mounting head. Using the pre-compensated mounting parameters as the driving command, the lifting drive shaft of the mounting head is driven to descend. During the descent of the mounting head, the current pressure feedback value of the mounting head is captured in real time. The current pressure feedback value is separated from the preset target pressure value, and the deviation obtained is converted into the driving voltage adjustment increment of the mounting head. The incremental adjustment of the driving voltage is periodically fed back and superimposed on the driving voltage command of the mounting head until the current pressure feedback value reaches the target pressure value.

10. A closed-loop control system for placement pressure in a pick-and-place machine based on multi-sensor fusion, characterized in that, The system is used to implement the multi-sensor fusion-based closed-loop control method for placement pressure in a pick-and-place machine as described in claim 1, the system comprising: The signal acquisition module is used to acquire the vacuum pressure fundamental frequency and real-time speed of the placement head in the target placement machine, and to synchronously cut the uniform descent segment of the vacuum pressure fundamental frequency to obtain the vacuum pressure subsequence and vacuum pressure baseline value of the placement head. The normalized pressure drop analysis module is used to perform normalized pressure drop analysis on the vacuum pressure subsequence with the vacuum pressure baseline value as a reference, so as to obtain the approximate pressure drop sequence of the mounting head. The response feature extraction module is used to perform first-order inertial element time-domain feature fitting on the proximity pressure drop sequence to obtain the proximity response time constant and proximity response steady-state amplitude of the mounting head. The gap height mapping module is used to map the steady-state amplitude of the proximity response to the gap height to obtain the minimum gap height between the component carried in the mounting head and the target substrate. The skew angle inversion module is used to perform aerodynamic impedance inversion on the lower surface of the carrying element based on the proximity response time constant and the proximity response steady-state amplitude, so as to obtain the attitude skew angle of the lower surface of the carrying element. The pre-compensation and pressure closed-loop module is used to perform pre-compensation of the contact posture of the mounting head based on the lower surface attitude skew angle and the minimum gap height, so as to obtain the mounting stroke parameters of the mounting head.

Images