An ultra-high molecular weight polyethylene lined tubing processing control system and method

Abstract

The present application relates to the technical field of program control system, specifically to a kind of super high molecular polyethylene lining oil pipe processing control system and method, system includes: process parameter solution module, obtains the input steel pipe outer diameter, lining pipe wall thickness, material thermal expansion coefficient and elastic modulus parameter.In the present application, in the processing process, by obtaining steel pipe outer diameter, lining pipe wall thickness, material thermal expansion coefficient and elastic modulus parameter and the like basic data and the heating temperature and pressure-holding pressure of each time point are combined, so that control command can match process demand in advance, so as to realize accurate response when driving heating and pressurizing device in execution phase.Synchronous start time record and sensor data acquisition, the real-time values of temperature and pressure are associated with time stamp and integrated into complete process curve, so that process information has continuity and traceability.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of program control system, and particularly relates to an ultra-high molecular polyethylene inner lining oil pipe processing control system and method. BACKGROUND

[0002] The technical field of program control system covers the related technologies of using programmable instructions and logic flow to automatically control equipment, processes and production processes. The core of this field is to combine the data collected by sensors with the preset control logic through software and hardware collaboration to drive the actuator to achieve precise operation and state adjustment.

[0003] In actual application, the prior art mostly relies on preset control logic to drive the actuator, which makes it difficult for the control instructions to fully match different materials and structural conditions, affecting the accuracy of heating and pressurization. Data collection is often limited to single-point recording or periodic sampling, making it difficult to accurately reflect the subtle fluctuations in the processing process. In terms of trajectory evaluation, it is not sensitive to slight but continuous deviation changes. Therefore, improvement is needed. SUMMARY

[0004] The purpose of the present application is to solve the shortcomings in the prior art and propose an ultra-high molecular polyethylene inner lining oil pipe processing control system and method.

[0005] In order to achieve the above-mentioned purpose, the present application adopts the following technical scheme: an ultra-high molecular polyethylene inner lining oil pipe processing control system comprises:

[0006] A process parameter solving module obtains the input steel pipe outer diameter, inner lining pipe wall thickness, material thermal expansion coefficient and elastic modulus parameters, and establishes online feedforward control instructions;

[0007] A processing process execution module calls the online feedforward control instructions, drives the heating and pressurizing device, synchronously starts time recording and sensor data collection, obtains real-time processing process data stream, collects and integrates the temperature sensor and pressure sensor values in the real-time processing process data stream, and associates them with time stamps to generate this batch process curve data;

[0008] A trajectory consistency evaluation module extracts the target time sequence data in the online feedforward control instructions and the time sequence data in the this batch process curve data, performs nonlinear alignment path search, establishes a normalized path deviation set, calculates the normalized path deviation set, and obtains a process trajectory consistency score;

[0009] A processing stability judgment module compares the process trajectory consistency score with a preset processing quality judgment threshold to obtain a consistency judgment state, and generates a processing process stability alarm according to the consistency judgment state.

[0010] Preferably, the online feedforward control instruction obtaining step is:

[0011] Based on the outer diameter of the steel pipe, the wall thickness of the lining pipe, the material thermal expansion coefficient and the elastic modulus parameters, a time increasing sequence is established, the target heating temperature and the target holding pressure caused by thermal expansion are calculated point by point, and the target parameters at each time point are recorded with a unified timestamp to form a heating temperature and holding pressure combination;

[0012] According to the heating temperature and holding pressure combination, the heating temperature and holding pressure at each time point are mapped to the parameter bits and written into the timestamp and the check field according to the device instruction byte sequence and the field boundary, a complete frame is spliced by using a fixed message header and a message tail, and a device executable instruction format is formed;

[0013] According to the device executable instruction format, the batch identification and the pipe parameter are bound, the execution beat and the allowed error range are set, the channel priority and the start-stop mark are configured, and the feedback code is used for state confirmation, the integrity check and the preparation for issuing are completed, and the online feedforward control instruction is generated.

[0014] Preferably, the real-time processing process data stream obtaining step is:

[0015] Based on the online feedforward control instruction, the field sequence and the check field are analyzed and the heating channel setting value and the pressurizing channel setting value are extracted, written into the heating and pressurizing device control register and set the start mark, write the unified timestamp in the time record and enable the sensor data acquisition, form the online feedforward control instruction execution sequence;

[0016] Based on the online feedforward control instruction execution sequence, the temperature sensor and the pressure sensor values are triggered to read according to the unified timestamp, the unified timestamp and the temperature sensor and the pressure sensor values are packaged into a record frame according to the fixed field sequence and appended in time sequence, and a real-time processing process data stream is formed.

[0017] Preferably, the batch process curve data obtaining step is:

[0018] Based on the real-time processing process data stream, the unified timestamp and the temperature sensor and the pressure sensor values are extracted frame by frame and sorted according to the unified timestamp, the records with the same unified timestamp are merged, and after deleting the repeated frames and the missing field frames, the curve segments are continuously spliced to generate the batch process curve data.

[0019] Preferably, the regularized path deviation set obtaining step is:

[0020] Based on the online feedforward control instruction and the batch process curve data, the target timestamp and the target value in each target timestamp in the online feedforward control instruction are analyzed, the target timestamp and the target value are combined into a target time sequence, the corresponding timestamp and the sensor value in the batch process curve data are extracted and combined into a time sequence, and the target time sequence data and the time sequence data are formed;

[0021] According to the target time sequence data and the time sequence data, the start index and the end index are set and the time index is incremented, the Euclidean distance of the two-dimensional vector composed of the temperature deviation and the pressure deviation of the target time sequence data and the time sequence data is taken as the matching cost in the three-neighborhood candidate, the pairing index is recorded along the path with the minimum cumulative cost, and the normalized temperature deviation and the normalized pressure deviation are calculated point by point to form a set of normalized path deviations.

[0022] Preferably, the process of obtaining the process trajectory consistency score comprises the following steps:

[0023] According to the set of normalized path deviations, the process trajectory consistency score is calculated.

[0024] Preferably, the process of obtaining the consistency determination state comprises the following steps:

[0025] Based on the process trajectory consistency score and the preset processing quality determination threshold, the numerical format of the process trajectory consistency score is analyzed and the validity is confirmed, the preset processing quality determination threshold is called and the numerical precision is unified, the size comparison is performed under the same numerical precision, and the consistency comparison result is generated;

[0026] According to the consistency comparison result, it is determined whether the process trajectory consistency score is within or beyond the preset processing quality determination threshold range, the determination result is recorded and the state identifier is assigned, the state identifier and the determination timestamp are bound, and the consistency determination state is generated.

[0027] Preferably, the process of obtaining the processing process stability alarm comprises the following steps:

[0028] Based on the consistency determination state, the state identifier corresponding to the consistency determination state is retrieved and the determination range information is analyzed, if the consistency determination state identifier indicates that the preset processing quality determination threshold range is exceeded, the alarm generation task is written in the event trigger queue and the alarm type and the trigger time are bound, and the processing process stability alarm is generated.

[0029] The application also provides a processing control method for an ultra-high molecular polyethylene inner lining oil pipe, which comprises the following steps:

[0030] An online feedforward control instruction is obtained by inputting the outer diameter of the steel pipe, the inner lining pipe wall thickness, the material thermal expansion coefficient and the elastic modulus parameters;

[0031] calling the online feedforward control instruction to drive the heating and pressurizing device, synchronously starting time recording and sensor data acquisition, obtaining a real-time processing process data stream, collecting and integrating temperature sensor and pressure sensor values in the real-time processing process data stream, and associating the values with time stamps to generate batch process curve data;

[0032] extracting target time sequence data in the online feedforward control instruction and time sequence data in the batch process curve data, performing nonlinear alignment path searching, establishing a set of normalized path deviations, calculating the set of normalized path deviations to obtain a process trajectory consistency score;

[0033] comparing the process trajectory consistency score with a preset processing quality determination threshold to obtain a consistency determination state, and generating a processing process stability alarm according to the consistency determination state.

[0034] Compared with the prior art, the advantages and positive effects of the present application are that:

[0035] In the present application, in the processing process, by obtaining basic data such as the outer diameter of the steel pipe, the wall thickness of the inner liner pipe, the material thermal expansion coefficient and the elastic modulus parameter, and combining the heating temperature and the holding pressure at each time point, the control instruction can match the process requirements in advance, so that precise response can be realized when driving the heating and pressurizing device in the execution stage. Synchronous starting of time recording and sensor data acquisition, association and integration of real-time values of temperature and pressure with time stamps into complete process curves make the process information have continuity and traceability. By extracting target time sequence data and actual time sequence data, performing nonlinear alignment path searching and constructing a set of normalized path deviations, and then accumulating the differences in the deviation set for calculation, the trajectory comparison not only reflects the static difference, but also reflects the overall consistency of the processing process. Comparing the process trajectory consistency score with a preset processing quality determination threshold, the determination state obtained can trigger a stability alarm in time when the control standard is exceeded, so that abnormal processing trends can be discovered and intervened in advance. The closed-loop control from parameter calculation to process monitoring, trajectory comparison and quality warning is realized, and the stability of the processing process, the accuracy of the data and the real-time of the quality determination are improved. BRIEF DESCRIPTION OF DRAWINGS

[0036] Figure 1 The system flowchart of the present application. DETAILED DESCRIPTION

[0037] In order to make the purpose, technical scheme and advantages of the present application clearer, the present application will be further described in detail below in combination with the drawings and examples. It should be understood that the specific examples described herein are only used to explain the present application and do not limit the present application.

[0038] Please refer toFigure 1 The application provides a technical solution: an ultra-high molecular weight polyethylene inner tube processing control system comprising:

[0039] A process parameter solving module obtains the input steel pipe outer diameter, inner lining pipe wall thickness, material thermal expansion coefficient and elastic modulus parameters, and establishes an online feedforward control instruction.

[0040] A processing process execution module calls the online feedforward control instruction, drives the heating and pressurizing device, synchronously starts time recording and sensor data acquisition, obtains real-time processing process data flow, collects and integrates the temperature sensor and pressure sensor values in the real-time processing process data flow, and associates them with time stamps to generate this batch process curve data.

[0041] A trajectory consistency evaluation module extracts the target time sequence data in the online feedforward control instruction and the time sequence data in the this batch process curve data, performs nonlinear alignment path searching, establishes a regularized path deviation set, calculates the regularized path deviation set, and obtains a process trajectory consistency score.

[0042] A processing stability judgment module compares the process trajectory consistency score with a preset processing quality judgment threshold to obtain a consistency judgment state, and generates a processing process stability alarm according to the consistency judgment state.

[0043] The acquisition step of the online feedforward control instruction is:

[0044] Based on the steel pipe outer diameter, inner lining pipe wall thickness, material thermal expansion coefficient and elastic modulus parameters, a time increment sequence is established, the target heating temperature and target pressure maintaining pressure caused by thermal expansion are calculated point by point, and the target parameters at each time point are recorded with a unified time stamp to form a heating temperature and pressure maintaining pressure combination.

[0045] According to the heating temperature and pressure maintaining pressure combination, the heating temperature and pressure maintaining pressure at each time point are mapped to the parameter bits and written into the time stamp and the check field according to the device instruction byte order and field boundary, a fixed message header and message tail are spliced to complete the frame, and a device executable instruction format is formed.

[0046] According to the device executable instruction format, the batch identification and pipe material parameters are bound, the execution beat and allowable error range are set, the channel priority and start-stop markers are configured, and the feedback code is used for state confirmation, the integrity check and preparation for issuing are completed, and the online feedforward control instruction is generated.

[0047] Specifically, based on the steel pipe outer diameter, inner lining pipe wall thickness, material thermal expansion coefficient and elastic modulus parameters, the entire processing process is first discretized in the time dimension, the total processing time is set, for example, 600 seconds according to historical production data, and a basic time step of 1 second is set, an array of 600 elements is established, and the first element is set as the initial time point, and the last element is set as the final time point. seconds to seconds, a time-increasing sequence containing 601 time points in total, then, for each time point in the sequence , the target parameter is calculated point by point according to the thermal coupling model, the specific calculation process of the thermal coupling model is that, first, the target radial expansion amount that the liner pipe needs to reach at time point is determined based on the thermal physical properties of the material, which needs to ensure that the outer wall of the liner pipe and the inner wall of the steel pipe are closely fitted, the calculation considers not only the linear thermal expansion of the liner pipe itself, that is, the increment of the outer diameter of the liner pipe is equal to the product of the initial outer diameter, the thermal expansion coefficient of the material and the temperature difference, but also the synchronous expansion of the steel pipe inner wall after heating, by solving the contact interference amount of the two at a specific temperature to deduce the required target heating temperature, at the same time, in order to ensure that the liner pipe always maintains effective contact with the inner wall of the steel pipe during the heating expansion process and overcomes the surface unevenness, a dynamic target holding pressure needs to be applied, the calculation of the pressure value is based on the change of the elastic modulus of the material at different temperatures, by establishing a functional relationship between pressure and temperature, for example, a linear relationship , wherein is the initial contact reference pressure, which is set according to historical data statistics according to the pipe specifications, for example, it is set to 0.1 MPa for a specific pipe specification, k is the pressure-temperature coefficient, which reflects the influence of material softening on the required pressure, which is fitted according to material test data, for example, it is set to 0.01 MPa / ℃, is the target heating temperature calculated at the current time point, is the initial ambient temperature, is the target holding pressure calculated at time point ti, the target heating temperature and the target holding pressure calculated at each time point are bound with a unified format timestamp, for example, the millisecond number since the process starts, to form a three-tuple data containing the timestamp, the target temperature and the target pressure, finally, the three-tuple data calculated at all time points (from to ) are arranged in chronological order to form a combination of heating temperature and holding pressure.

[0048] According to the heating temperature and pressure combination, a binary data frame conforming to the communication protocol of the lower machine controller is constructed for each data triplet at a time point. The process first determines the byte order and field boundary of the device instruction, for example, defines a fixed data frame structure with a total length of 18 bytes, where the byte order is big-endian mode, and the field boundary is divided as follows: byte 0 is the fixed message header, set to 0xAA, byte 1 is the data frame length field, fixed at 16, bytes 2-5 are a 32-bit unsigned integer for storing timestamp data in milliseconds, bytes 6-9 are a 32-bit floating point number following the IEEE 754 standard for storing target heating temperature values, bytes 10-13 are another 32-bit floating point number also following the IEEE 754 standard for storing target pressure values, bytes 14-15 are a 16-bit cyclic redundancy check code for data verification, byte 17 is the fixed message tail, set to 0x55. Then, each triplet in the heating temperature and pressure combination is traversed, for example, for a triplet with a timestamp of 2000 milliseconds, a target heating temperature of 125.5°C, and a target pressure of 1.5 MPa, first convert the timestamp 2000 to hexadecimal 0x000007D0, convert the temperature 125.5 to its IEEE 754 single-precision floating point representation 0x42FB0000, and convert the pressure 1.5 to 0x3FC00000. Then, according to the preset CRC-16 / MODBUS algorithm, calculate all bytes from the data frame length field (byte 1) to the pressure data field (byte 13) to generate a 16-bit check field. All byte data obtained by the above conversion and calculation are spliced in the pre-defined order, i.e. the message header, data length, timestamp, temperature, pressure, check field, and finally the message tail is spliced to complete the construction of a complete frame. Repeat this process for all data points in the heating temperature and pressure combination to generate a series of complete frames arranged in time sequence to form a device executable instruction format.

[0049] According to the instruction format executable by the device, final configuration and packaging are performed before delivery. First, a globally unique batch identifier is bound to the instruction set of the batch, which is composed of the production date, device number and serial number of the day, for example "20240520-SN001-034". At the same time, the outer diameter of the steel pipe, the wall thickness of the inner liner, the material thermal expansion coefficient and the elastic modulus parameters used in this processing are attached to the instruction header as metadata for production process tracing. Then, the execution parameters of the instructions are set, where the execution beat is set according to the time step of the previous generation instruction, for example, if the time step is 1 second, the execution beat is set to 1 Hz, ensuring that the controller sends a frame of instructions every second, and the allowable error range needs to be set according to the process requirements. This setting process is based on statistical analysis of historical qualified product processing data. For example, statistical data shows that products with a temperature deviation of no more than 1.5% and a pressure deviation of no more than 3% have a final bonding strength in the excellent grade. Therefore, the allowable error of temperature is set to ±1.5% of the target value, and the allowable error of pressure is set to ±3.0% of the target value. A specific example is given to illustrate that when the target temperature is 200℃, the allowable actual temperature range is 197℃ to 203℃, and when the target pressure is 1.0MPa, the allowable actual pressure range is 0.97MPa to 1.03MPa. Then, the parameters for the coordinated work of multiple channels are configured. Since the response speed of the pressure system is usually faster than that of the temperature system, the priority of the pressure control channel is set to high (for example, the value is 1), and the priority of the temperature control channel is set to standard (for example, the value is 10), to ensure that the pressure can quickly follow the instruction changes. At the same time, a specific start marker byte, such as 0x01, is placed in the first frame of the instruction sequence, and an end marker byte 0x00 is appended after the last frame, which is used to control the start and stop of the device. A receipt code mapping table is associated with each instruction, which defines the possible returned states after device execution, such as 0x00 representing "instruction received", 0x01 representing "instruction execution success", and 0xFF representing "instruction verification error". Finally, a complete data packet containing all instruction frames, batch identifier, pipe parameters and configuration parameters is subjected to a whole SHA-256 hash calculation to generate a final checksum, which is appended to the end of the data packet to complete the integrity check and preparation for delivery, generating online feedforward control instructions.

[0050] The real-time processing process data flow acquisition step is:

[0051] Based on the online feedforward control instructions, the field order and verification field are analyzed and the heating channel set value and the pressurizing channel set value are extracted, written into the heating and pressurizing device control register and the start marker is set, the unified timestamp is written into the time record and the sensor data acquisition is enabled, forming an online feedforward control instruction execution sequence;

[0052] Based on the online feedforward control instruction execution sequence, the temperature sensor and pressure sensor values are triggered to read at a unified timestamp, and the unified timestamp and temperature sensor and pressure sensor values are packaged as a record frame in a fixed field order and appended in time sequence to form a real-time processing process data stream.

[0053] Specifically, based on the online feedforward control instruction, the system controller first reads the instruction data from the communication queue frame by frame. For each frame of data, first locate byte 0 and byte 17, check if the message header is 0xAA and the message tail is 0x55, if the check is not passed, discard the frame and record the error log, if it is passed, according to the check field of byte 14 to 15, using the preset CRC-16 / MODBUS algorithm to calculate the data content of byte 1 to 13, and compare the calculation result with the value of the check field, only when the two are completely consistent, it is considered that the instruction frame is valid, then, according to the pre-defined field order, extract data from the valid instruction frame, specifically, read 32-bit data from byte 6 to 9 and parse it into a single-precision floating-point number conforming to IEEE754 standard, get the heating channel set value, for example, read the hexadecimal value 0x43480000, which is parsed as 200.0℃, similarly, read and parse from byte 10 to 13 to get the pressure channel set value, for example, read 0x3F800000, which is parsed as 1.0MPa, then, write the parsed heating channel set value to the control register address associated with the heating device, such as the D100 data register of PLC, and write the pressure channel set value to the control register address associated with the pressure device, such as the D200 data register, after the value is written, immediately write a set signal "1" to a specific start control register, such as M10, this operation triggers the heating and pressure device to adjust the power and pressure according to the new set value, at the same time, the system obtains the current high-precision system timestamp, such as the millisecond count since the process starts, and writes this timestamp into the internal time record log, and immediately activates the data acquisition channel connected with the temperature sensor and pressure sensor, sets its acquisition frequency, which is set according to historical data analysis, generally 5 to 10 times of the control frequency to capture the dynamic process, for example, if the control beat is 1Hz, the acquisition frequency is set to 10Hz, forming the online feedforward control instruction execution sequence.

[0054] Based on the online feedforward control instruction execution sequence, the internal timer of the system generates an accurate trigger interrupt every 100 milliseconds according to the preset 10Hz acquisition frequency, which synchronously starts the reading process of the temperature sensor and pressure sensor values. At each interrupt trigger, the system first sends a read instruction to the temperature sensor front-end analog-to-digital converter through the SPI or I2C bus to obtain the raw digital quantity representing the current temperature, and according to the calibration curve of the sensor, such as a third-order polynomial where V is the read voltage signal, a, b, c, d are pre-calibrated coefficients, T represents the temperature, converting the digital quantity into a physical value in Celsius units, at the same time, through another independent channel, a read instruction is sent to the signal conditioning circuit of the pressure sensor, to obtain the original digital quantity representing the current pressure, and a linear conversion formula is applied where N is the original digital quantity, P represents the pressure, k is a proportional coefficient determined according to the pressure calibration experiment, for example, 0.001 MPa / LSB, is the zero-point offset, which is converted into a physical value in megapascal units, after obtaining the two physical values of temperature and pressure, a unified timestamp associated with the current interrupt is immediately obtained, then, the three data items are encapsulated in a fixed field order to construct a 12-byte record frame, the structure of which is defined as follows: bytes 0 to 3 store the unified timestamp in 32-bit unsigned integer format, bytes 4 to 7 store the temperature sensor value in 32-bit floating-point format, bytes 8 to 11 store the pressure sensor value in 32-bit floating-point format, finally, the newly generated record frame is appended to a first-in-first-out (FIFO) memory buffer, over time, this buffer will continuously accumulate these record frames generated in chronological order, forming a real-time processing process data stream.

[0055] The acquisition steps of the batch process curve data are as follows:

[0056] Based on the real-time processing process data stream, the unified timestamp and the temperature sensor and pressure sensor values are extracted frame by frame and sorted by the unified timestamp, the records with the same unified timestamp are merged and the duplicate frames and missing field frames are deleted, then the continuous splicing is performed to generate the batch process curve data.

[0057] Specifically, based on the real-time processing process data stream, the system starts the data post-processing process after completing a complete batch of processing or after reaching the preset data buffer size. The process first traverses all the record frames in the buffer. For each frame, according to the preset 12-byte frame structure, the uniform timestamp is extracted from bytes 0 to 3, the temperature sensor value is extracted from bytes 4 to 7, and the pressure sensor value is extracted from bytes 8 to 11. The extracted three-tuple data (timestamp, temperature, pressure) is temporarily stored in a temporary data structure. After all the frames are extracted, all the records in the temporary data structure are sorted in ascending order with the uniform timestamp as the sorting key to correct the slight disorder caused by multi-threading or interruption delay. After sorting, the system checks whether there are records with the same uniform timestamp. If there are, the arithmetic mean of the temperature sensor values and the pressure sensor values of these records is calculated, a new merged record is generated using the average value, and the original multiple repeated timestamp records are deleted. Then, the data is checked for validity, and incomplete records are deleted one by one, such as repeated frames with invalid values (such as NaN) in the temperature or pressure field due to communication errors. At the same time, the system also deletes obviously abnormal data points according to the preset physical rationality range, which is set according to the device safety operation procedures and material physical properties. For example, the temperature is compared with the interval of 0℃ to 300℃, and the pressure is compared with the interval of 0MPa to 3MPa. Any record outside this interval is considered as a wild point generated by sensor instantaneous failure or strong interference and is removed. After completing the above merging and deleting operations, all the remaining valid data points in chronological order are concatenated to form a series of discrete point sets representing the changes of temperature and pressure over time, generating the batch process curve data.

[0058] The acquisition step of the normalized path deviation set is:

[0059] Based on the online feedforward control instruction and the batch process curve data, each target timestamp and target value in the online feedforward control instruction are analyzed, the target timestamp and target value are combined into a target time sequence, the corresponding timestamp and sensor value in the batch process curve data are extracted, and combined into a time sequence, forming target time sequence data and time sequence data;

[0060] According to the target time sequence data and the time sequence data, the start index and the end index are set and the time index is incremented. In the three-neighborhood candidate, the Euclidean distance of the two-dimensional vector composed of the temperature deviation and the pressure deviation of the target time sequence data and the time sequence data is taken as the matching cost, the pairing index is recorded along the path with the minimum cumulative cost, the normalized temperature deviation and the normalized pressure deviation are calculated point by point, and the normalized path deviation set is formed.

[0061] Specifically, based on the online feedforward control instruction and the batch process curve data, first, the online feedforward control instruction is analyzed frame by frame. The specific operation is that the system reads the first frame of 18 bytes of data from the starting position of the instruction sequence, checks whether the message header (byte 0) and the message tail (byte 17) are respectively 0xAA and 0x55, and uses the CRC-16 / MODBUS algorithm to check and calculate the data content (bytes 1 to 13). After comparing the check field (bytes 14 to 15) and confirming that there is no error, the data is extracted according to the field boundary definition, that is, 32-bit unsigned integers from bytes 2 to 5 are extracted as target time stamps, 32-bit floating-point numbers from bytes 6 to 9 are extracted as target heating temperatures, and 32-bit floating-point numbers from bytes 10 to 13 are extracted as target holding pressures. The three (target time stamp, target heating temperature, target holding pressure) are stored as a data unit in a newly created sequence, then the next frame of instructions is processed, and the process is repeated until all instruction frames are parsed. Finally, a target time sequence sequence arranged in time sequence is formed. At the same time, the system directly loads the generated batch process curve data, which is itself a sensor value set sorted by time stamp. The system reads the data set into memory as a whole and constructs a data structure containing uniform time stamps, temperature sensor values and pressure sensor values. This data structure is the time sequence, and finally the target time sequence data and the time sequence data are formed.

[0062] According to the target time sequence data and the time sequence data, a variant of dynamic time warping algorithm is used for nonlinear path alignment. First, a two-dimensional cost matrix is initialized, the number of rows is equal to the length of the target time sequence data, the number of columns is equal to the length of the time sequence data, and the starting point index is set to the top left corner of the matrix (0, 0), and the ending point index is set to the bottom right corner. Then, fill in the cost matrix. For any point (i, j) in the matrix, the local matching cost is determined by calculating the Euclidean distance between the temperature and pressure values of the i th point in the target time sequence data and the temperature and pressure values of the j th point in the time sequence data. Specifically, calculate the temperature deviation (target temperature i - actual temperature j) and pressure deviation (target pressure i - actual pressure j), then take the square sum of the two deviation values and take the square root to get the matching cost of the point. During the filling process, the cumulative cost of each point is the local matching cost of itself plus the minimum cumulative cost value of its left (point (i, j-1)), upper (point (i-1, j)) and upper left (point (i-1, j-1)) three adjacent points. This rule naturally limits the monotonic increase of time index. After the filling of the entire cost matrix is completed, start from the ending point index and backtrack to the starting point index along the direction with the smallest cumulative cost to record the index pairs of all points on the path to form a paired index sequence. Here, each "index pair" in the paired index sequence defines a corresponding relationship between the target sequence and the actual sequence. The data point pair determined by the index pair is called an "alignment point" or a "paired point". In the following, "alignment point" and "paired point" have the same meaning. Then, traverse the paired index sequence. For each alignment point represented by an index pair (i, j), calculate its actual temperature deviation (target temperature i - actual temperature j) and pressure deviation (target pressure i - actual pressure j), and then normalize these deviation values. The temperature normalization constant and the pressure normalization constant used for normalization are determined according to the maximum allowable tolerance in the process specification. For example, the process document specifies that during the steady-state processing phase, the temperature fluctuation should not exceed ±5℃, and the pressure fluctuation should not exceed ±0.2MPa. Therefore, the temperature normalization constant is 5.0, and the pressure normalization constant is 0.2. Divide the temperature deviation of each paired point by 5.0 to get the normalized temperature deviation, and divide the pressure deviation by 0.2 to get the normalized pressure deviation. Finally, combine the normalized temperature deviations and normalized pressure deviations of all alignment points to form a set of normalized path deviations.

[0063] The process trajectory consistency score is obtained by:

[0064] According to the set of normalized path deviations, the process trajectory consistency score is calculated, and the calculation formula is:

[0065] ;

[0066] wherein, R is the number of alignment points, is the dimensionless value of the temperature deviation of the rth alignment point divided by the temperature normalization constant, is the dimensionless value of the pressure deviation of the rth alignment point divided by the pressure normalization constant, is the dimensionless value of the temperature deviation of the rth alignment point divided by the temperature normalization constant, is the dimensionless value of the pressure deviation of the rth alignment point divided by the pressure normalization constant, is the dimensionless value of the temperature deviation of the rth alignment point divided by the temperature normalization constant, is the dimensionless value of the pressure deviation of the rth alignment point divided by the pressure normalization constant, is the dimensionless dynamic influence coefficient, used to adjust the proportion of the deviation jump amplitude in the total score.

[0067] Specifically, the formula is: The advantage of the formula is that the static deviation amplitude of the actual process curve and the target setting curve at each alignment time point is quantified, that is, it is embodied by the Euclidean distance A dynamic penalty term is introduced, which is specially used to punish the violent fluctuations and instability in the process, that is, the rapid change of the deviation. The introduction of the parameter makes it possible to flexibly adjust the sensitivity to process volatility according to the requirements of different processes for stability, so that the final consistency score can more comprehensively and accurately reflect the processing quality, avoiding one-sided evaluation by only considering the average deviation size while ignoring process stability;

[0068] R is the number of alignment points, which is obtained by searching the nonlinear alignment path of the target time series data and the time series data in the previous step. It represents the total number of paired data points contained in the finally determined optimal alignment path. For example, in the dynamic time warping alignment process described above, the algorithm backtracks to find a path containing 650 index pairs, so the value of R is 650.

[0069] and are the normalized temperature deviation and the normalized pressure deviation of the rth alignment point, respectively. These two parameters are also directly derived from the regularization path deviation set generated in the previous step. The acquisition process is as follows: first, calculate the original deviation between the target time series data and the time series data at the rth alignment point, that is, and , where is the original temperature deviation, is the original pressure deviation, is the target heating temperature value corresponding to the rth alignment point, is the actual temperature sensor value corresponding to the alignment point, the target pressure value corresponding to the alignment point, the actual pressure sensor value corresponding to the alignment point, and then normalized using a normalization constant, which is set according to the quality control standard document of the production process of the oil pipe of this model, which clearly specifies the maximum allowable process deviation at each stage, for example, for a certain specification of oil pipe, the temperature deviation of the heating stage is allowed to be up to 5℃, and the pressure deviation of the pressure holding stage is allowed to be up to 0.25MPa, therefore, the temperature normalization constant is 5.0, and the pressure normalization constant is 0.25, so the calculation formula is: and for example, at the 100th alignment point (P100) , the calculated original temperature deviation is -2.0℃, and the original pressure deviation is 0.05MPa, so the corresponding dimensionless value is: , ;

[0070] is the dimensionless dynamic influence coefficient, which is used to adjust the weight of the deviation jump amplitude in the total score, and its value is determined based on statistical analysis and optimization of historical production data, the specific steps are as follows, first, collect the regularization path deviation set data of at least 100 completed production batches and the final product quality inspection report (for example, the test results of the bonding strength between the liner and the steel pipe) corresponding to these batches, for each batch, calculate the cumulative sum of the static deviation term and the cumulative sum of the dynamic deviation term , at the same time, quantify the bonding strength data in the product quality inspection report, normalize the strength value to 0-100 points as the quality score Q, then build an optimization objective function, that is, find a value, so that the absolute value of the Pearson correlation coefficient of the score sequence calculated by and the quality score Q sequence is maximum, by traversing the value in the preset interval (for example, 0.1 to 5.0, step 0.1), calculate the correlation coefficient corresponding to each value, finally select the value with the strongest correlation, for example, it is found through calculation that when , has the strongest negative correlation with Q (the correlation coefficient is -0.92), indicating that the score at this time can best reflect the quality of the product, so the value is set to be ;

[0071] Calculation process:

[0072] According to the foregoing parameter acquisition steps, the specific example parameters are set as follows: , , and the data of the three consecutive alignment points from the set of normalized path deviations are taken out:

[0073] The r-1th (e.g., r = 99) alignment point: , ;

[0074] The rth (e.g., r = 100) alignment point: , ;

[0075] The r+1th (e.g., r = 101) alignment point: , ;

[0076] The value of the rth item is calculated from , which is taken as an example to calculate the 100th item ( ) and the 101th item ( ):

[0077] When , the value of the item is calculated:

[0078] The static deviation item: ;

[0079] The dynamic deviation item: ;

[0080] The value of the 100th item = ;

[0081] When , the value of the item is calculated:

[0082] The static deviation item: ;

[0083] The dynamic deviation item: ;

[0084] The value of the 101th item = ;

[0085] The final process trajectory consistency score is obtained by accumulating and summing all the calculated items from to :

[0086] ;

[0087] For example, after all the calculations, the final result is .

[0088] The result shows that the consistency score of the batch processing process is 315.8, which comprehensively reflects the degree of deviation of the actual process parameters from the target value and the stability of the process during the entire processing process. The lower the score, the better the consistency and the higher the processing quality.

[0089] The consistency determination state obtaining step is:

[0090] Based on the process trajectory consistency score and the preset processing quality determination threshold, the numerical format of the process trajectory consistency score is analyzed and its validity is confirmed, the preset processing quality determination threshold is called and the numerical accuracy is unified, the size comparison is performed under the same numerical accuracy, and a consistency comparison result is generated;

[0091] According to the consistency comparison result, it is determined whether the process trajectory consistency score is within the preset processing quality determination threshold range or exceeds the preset processing quality determination threshold range, the determination result is recorded and a state identifier is assigned, the state identifier is bound with a determination time stamp, and a consistency determination state is generated.

[0092] Specifically, based on the process trajectory consistency score and the preset processing quality determination threshold, the system first checks the numerical format of the received process trajectory consistency score, for example, confirms that it is a standard double-precision floating-point number, and checks whether its numerical range falls within the expected interval, for example, greater than or equal to zero. If the format or range is invalid, an error is recorded and subsequent processing is terminated. After confirming the validity, the system calls the preset processing quality determination threshold corresponding to the current pipe specification and process type from the configuration file. The threshold is an interval including the upper and lower limits, which is set based on statistical analysis of a large amount of historical production data. The specific method is as follows: collect the process trajectory consistency scores and their corresponding finished product destructive test results (such as joint strength) of all products of the same specification in the past year, divide the test results into “excellent”, “good”, “qualified”, and “unqualified” four levels, and then calculate the distribution of process trajectory consistency scores corresponding to each level. Then, the upper limit of the 95% confidence interval of the “qualified” level score distribution is taken as the upper limit of the determination threshold, for example, 350.0, and the upper limit of the 95% confidence interval of the “excellent” level score distribution is taken as another determination reference, for example, 150.0, thereby forming a multi-level determination standard. The system then unifies the process trajectory consistency score and the called threshold to the same numerical accuracy, for example, all converted to floating-point numbers with two decimal places, to avoid comparison errors due to precision problems. Finally, under the same numerical accuracy, the process trajectory consistency score, for example, 315.80, is directly compared with the preset processing quality determination threshold upper limit 350.00, and a consistency comparison result of Boolean type is generated, for example, because 315.80 is less than 350.00, the comparison result is true (true).

[0093] According to the consistency comparison result, the system enters the judgment process. If the consistency comparison result is true, that is, the process trajectory consistency score is less than or equal to the upper limit of the preset machining quality judgment threshold, it is determined that the score is within the preset machining quality judgment threshold range. Further, the system will compare with the more detailed level threshold, for example, compare the score 315.8 with the "excellent" level threshold 150.0 and the "good" level threshold 280.0 (the threshold is also obtained from historical data statistics, which is the upper limit of the 95% confidence interval of the "good" level score distribution). Since 315.8 is greater than 280.0, the system will record the final judgment result as "qualified" and assign it a predefined status identification code, for example, 0x02 represents "qualified". If the consistency comparison result is false, that is, the score exceeds the upper limit of the threshold, the judgment result is recorded as "unqualified" and assigned a status identification code 0x03. This status identification code mapping table (for example: 0x00 represents "not determined", 0x01 represents "excellent", 0x02 represents "qualified", 0x03 represents "unqualified") is pre-stored in the system configuration. After determining the judgment result and the corresponding status identification, the system immediately obtains the current high-precision system timestamp, for example, "2024-05-20 15:30:45.123", and binds the timestamp with the status identification code to form a data structure containing the judgment result, the status identification code and the judgment timestamp, generating a consistency judgment state.

[0094] The acquisition step of the machining process stability alarm is:

[0095] Based on the consistency judgment state, the corresponding state identification of the consistency judgment state is retrieved and the judgment range information is analyzed. If the consistency judgment state identification indicates that it exceeds the preset machining quality judgment threshold range, an alarm generation task is written in the event trigger queue and the alarm type and trigger time are bound, generating a machining process stability alarm.

[0096] Specifically, based on the consistency determination state, the system starts a background monitoring service that continuously polls newly generated consistency determination states. When a new state is detected, the system immediately retrieves the state identification code inside it, for example, retrieves the state identification as 0x03, and queries the pre-set state identification code mapping table to analyze the determination range information corresponding to the identification as "exceeding the pre-set machining quality determination threshold range", i.e., "unqualified". Once the determination result is confirmed as unqualified, the system immediately creates a new alarm generation task in the internal event trigger queue, and the data structure of the task contains multiple key fields. First, bind a specific alarm type, which is represented by a pre-set alarm code, for example, "W001", which is defined in the alarm code library as "machining process stability unqualified". Second, write the determination timestamp in the consistency determination state that led to this alarm, for example, "2024-05-20 15:30:45.123", as the trigger time of the alarm into the task. At the same time, the task will also bind related context information, including the batch identification of the current batch, the pipe material parameters, the calculated process trajectory consistency score, and the threshold upper limit exceeded. Finally, push this complete alarm generation task object to the tail of the event queue, and take it out from the queue by a special alarm processing thread to perform subsequent operations, such as popping up a highlighted warning box on the central monitoring interface, sending an alarm email to a designated mailbox, or sending a notification to the on-duty engineer through an SMS gateway, to generate a machining process stability alarm.

Claims

1. An ultra-high molecular weight polyethylene lined tubing processing control system, characterized by, The system comprises: a process parameter solving module, which acquires the input steel pipe outer diameter, inner liner pipe wall thickness, material thermal expansion coefficient and elastic modulus parameters, and establishes an online feedforward control instruction; a processing process execution module, which calls the online feedforward control instruction, drives the heating and pressurizing device, synchronously starts time recording and sensor data acquisition, acquires real-time processing process data stream, collects and integrates temperature sensor and pressure sensor values in the real-time processing process data stream, and associates them with time stamps to generate this batch process curve data; a trajectory consistency evaluation module, which extracts target time sequence data in the online feedforward control instruction and time sequence data in the this batch process curve data, performs nonlinear alignment path searching, establishes a normalized path deviation set, calculates the normalized path deviation set, and obtains a process trajectory consistency score; a processing stability judgment module, which compares the process trajectory consistency score with a preset processing quality judgment threshold to obtain a consistency judgment state, and generates a processing process stability alarm according to the consistency judgment state; the acquisition step of the online feedforward control instruction is: based on the steel pipe outer diameter, inner liner pipe wall thickness, material thermal expansion coefficient and elastic modulus parameters, a time increasing sequence is established, the target heating temperature and target pressure maintaining pressure caused by thermal expansion are calculated point by point, and the target parameters at each time point are recorded with a unified time stamp to form a heating temperature and pressure maintaining pressure combination; according to the heating temperature and pressure maintaining pressure combination, the heating temperature and pressure maintaining pressure at each time point are mapped to parameter bits and written into time stamps and check fields in the order of device instruction bytes and field boundaries, a complete frame is spliced by using a fixed message header and a message tail, and a device executable instruction format is formed; according to the device executable instruction format, batch identification and pipe parameters are bound, execution tempo and allowable error range are set, channel priority and start-stop markers are configured, and a feedback code is used for state confirmation, integrity verification and preparation for issuance are completed, and an online feedforward control instruction is generated.

2. The ultra-high molecular weight polyethylene lined tubing processing control system of claim 1, wherein, the acquisition step of the real-time processing process data stream is: based on the online feedforward control instruction, the field order and the check field are analyzed, and the heating channel setting value and the pressurizing channel setting value are extracted, written into the heating and pressurizing device control register, and the start marker is set, the unified time stamp is written in the time recording, and the sensor data acquisition is enabled, and an online feedforward control instruction execution sequence is formed; based on the online feedforward control instruction execution sequence, the temperature sensor and pressure sensor values are triggered to be read according to the unified time stamp, the unified time stamp and the temperature sensor and pressure sensor values are packaged into a record frame in a fixed field order and appended in time sequence, and a real-time processing process data stream is formed.

3. The ultra-high molecular weight polyethylene lined tubing processing control system of claim 1, wherein, the acquisition step of the this batch process curve data is: based on the real-time processing process data stream, the unified time stamp and the temperature sensor and pressure sensor values are extracted frame by frame and sorted by the unified time stamp, the records with the same unified time stamp are merged, and after the repeated frames and missing field frames are deleted, the curve segments are continuously spliced to generate this batch process curve data.

4. The ultra-high molecular weight polyethylene lined tubing processing control system of claim 1, wherein, the acquisition step of the normalized path deviation set is: Based on the online feedforward control instruction and the batch process curve data, each target timestamp and target value in the online feedforward control instruction are analyzed, the target timestamp and target value are combined into a target time sequence, a corresponding timestamp and sensor value in the batch process curve data are extracted and combined into a time sequence, and target time sequence data and time sequence data are formed; According to the target time sequence data and the time sequence data, a start index and an end index are set and a time index is incremented, a two-dimensional vector of temperature deviation and pressure deviation of the target time sequence data and the time sequence data is used as a matching cost in three-neighborhood candidates, a pairing index is recorded along a path with the minimum cumulative cost, normalized temperature deviation and normalized pressure deviation are calculated point by point, and a regularized path deviation set is formed.

5. The ultra-high molecular weight polyethylene lined tubing processing control system of claim 1, wherein, The process trajectory consistency score is obtained by: According to the regularized path deviation set, the process trajectory consistency score is calculated.

6. The ultra-high molecular weight polyethylene lined tubing processing control system of claim 1, wherein, The consistency determination state is obtained by: Based on the process trajectory consistency score and the preset processing quality determination threshold, the numerical format of the process trajectory consistency score is analyzed and its validity is confirmed, the preset processing quality determination threshold is called and the numerical accuracy is unified, a size comparison is performed under the same numerical accuracy, and a consistency comparison result is generated; According to the consistency comparison result, it is determined whether the process trajectory consistency score is within or beyond the preset processing quality determination threshold range, the determination result is recorded and a state identifier is assigned, the state identifier is bound with a determination timestamp, and a consistency determination state is generated.

7. The ultra-high molecular weight polyethylene lined tubing processing control system of claim 1, wherein, The processing process stability alarm is obtained by: Based on the consistency determination state, the state identifier corresponding to the consistency determination state is retrieved and the determination range information is analyzed. If the consistency determination state identifier indicates that the preset processing quality determination threshold range is exceeded, an alarm generation task is written in an event trigger queue and an alarm type and a trigger time are bound, and a processing process stability alarm is generated.

8. The ultra-high molecular weight polyethylene lined tubing process control system of any one of claims 1-7, wherein, The method comprises the following steps: Obtain the input steel pipe outer diameter, inner liner pipe wall thickness, material thermal expansion coefficient and elastic modulus parameters, and establish an online feedforward control instruction; Call the online feedforward control instruction to drive the heating and pressurizing device, start time recording and sensor data collection synchronously, obtain real-time processing process data stream, collect and integrate temperature sensor and pressure sensor values in the real-time processing process data stream, and associate them with timestamps to generate batch process curve data; Extract target time sequence data in the online feedforward control instruction and time sequence data in the batch process curve data, perform nonlinear alignment path search, establish a regularized path deviation set, calculate the regularized path deviation set, and obtain a process trajectory consistency score; Compare the process trajectory consistency score with the preset processing quality determination threshold to obtain a consistency determination state, and generate a processing process stability alarm according to the consistency determination state.

Images