Remote monitoring and communication system for CNC machine tools based on industrial Ethernet

Abstract

This invention belongs to the field of digital control technology for CNC machine tools, specifically providing a remote monitoring and communication system for CNC machine tools based on industrial Ethernet. It mainly constructs an interval sequence by acquiring instruction arrival timestamps, extracts features to generate a delay jitter evaluation vector, maps this vector to a reference interpolation period to generate a phase difference factor and a degradation index, and calculates the feed rate correction based on these. The machine tool buffer is then reconstructed using this correction and the degradation index to generate a dynamic buffer queue architecture. Finally, a prefetch sequence is extracted and pulses are adjusted to output control pulses for the drive motor. This invention effectively suppresses the microscopic impact of network fluctuations on the transmission chain from the control source, ensuring the smoothness of high-speed feed of the servo motor, effectively eliminating irreversible machining vibration marks caused by data cycle disorder, and improving the surface roughness and aerodynamic performance of precision aerospace parts.

Description

Technical Field

[0001] This invention belongs to the field of digital control technology for CNC machine tools, and specifically relates to a remote monitoring and communication system for CNC machine tools based on industrial Ethernet. Background Technology

[0002] In the remote distributed manufacturing of precision aerospace parts, spatial coordinate data is often continuously transmitted at high frequency via industrial Ethernet by a control server. The physical interpolation calculation logic executed by the underlying servo control system of the CNC machine tool at a fixed hardware clock frequency is used to drive the cutting tool to fit the design contour of the part for cutting.

[0003] However, the communication links in workshops are generally limited by the queuing and forwarding strategies of switches and the electromagnetic interference of equipment. This results in a high degree of uncertainty in the transmission delay of machining trajectory command data packets and microsecond-level random jitter when they reach the machine tool. Consequently, there is a significant physical phase difference between the interpolation clock of the CNC system and the arrival cycle of the network command, which disrupts the smooth data digestion rhythm. Under these conditions, the servo motor generates uneven distribution of interpolation pulses during feeding. The micro-impact at the shaft end acts on the cutting tool through the mechanical transmission chain, directly causing irreversible physical machining marks on the workpiece surface. This seriously affects the surface roughness index and aerodynamic service performance of precision parts.

[0004] Therefore, it is necessary to propose a remote monitoring and communication system for CNC machine tools based on industrial Ethernet to solve the above problems. Summary of the Invention

[0005] This invention provides a remote monitoring and communication system for CNC machine tools based on industrial Ethernet. It effectively solves the problem in the prior art where random jitter in the transmission delay of industrial Ethernet causes a phase difference between the interpolation clock and the instruction cycle, resulting in uneven pulses in the servo motor and consequently physical machining marks on the workpiece surface. It achieves the suppression of the microscopic impact of network fluctuations on the transmission chain from the control source, ensures the smoothness of high-speed feed of the servo motor, effectively eliminates irreversible machining marks caused by data cycle disorder, and improves the surface roughness index and aerodynamic service performance of aerospace precision parts.

[0006] To achieve the above objectives, the present invention adopts the following technical solution:

[0007] In a first aspect, the present invention provides a remote monitoring and communication method for CNC machine tools based on industrial Ethernet, comprising:

[0008] Obtain the set of physical arrival timestamps of continuously received processing trajectory instruction data packets from the industrial Ethernet, and construct an arrival time interval sequence based on the set of physical arrival timestamps.

[0009] The statistical distribution features of the arrival time interval sequence are extracted, and the statistical distribution features are used to construct a network delay jitter evaluation vector that characterizes the current communication link congestion state.

[0010] The network latency jitter evaluation vector is jointly mapped and analyzed with the preset CNC system reference interpolation period constant to generate the interpolation phase difference factor characterizing the degree of command arrival misalignment and the jitter degradation index characterizing the degree of network fluctuation.

[0011] Based on the interpolation phase difference factor and jitter degradation index, kinematic compensation calculations are performed to obtain the feed speed correction reference amount used to adjust the running trajectory of the servo motor.

[0012] Based on the feed rate correction reference and the jitter degradation index, the machine tool receiving buffer area is spatially reconstructed to obtain a dynamic buffer queue architecture that matches the network state.

[0013] The instruction prefetch sequence is extracted based on the dynamic buffer queue architecture, and the pulse adjustment of the instruction prefetch sequence is combined with the feed rate correction reference value to output the real-time interpolation control pulse for driving the servo motor.

[0014] Secondly, the present invention provides a remote monitoring and communication system for CNC machine tools based on industrial Ethernet, comprising:

[0015] The time stamp acquisition module is used to acquire the set of physical arrival timestamps of the processing trajectory instruction data packets continuously received by the industrial Ethernet, and to construct an arrival time interval sequence based on the set of physical arrival timestamps.

[0016] Feature extraction module: used to extract the statistical distribution features of the arrival time interval sequence, and use the statistical distribution features to construct a network delay jitter evaluation vector that characterizes the current communication link congestion state.

[0017] Link evaluation module: It is used to perform joint mapping analysis between the network latency jitter evaluation vector and the preset CNC system reference interpolation period constant, and generate the interpolation phase difference factor that characterizes the degree of command arrival misalignment and the jitter degradation index that characterizes the degree of network fluctuation.

[0018] Feed compensation module: used to perform kinematic compensation calculations based on interpolation phase difference factor and jitter degradation index to obtain the feed speed correction reference amount for adjusting the running trajectory of servo motor.

[0019] Cache reconstruction module: Based on the feed rate correction baseline and jitter degradation index, it performs spatial reconstruction of the machine tool receiving buffer area to obtain a dynamic cache queue architecture that matches the network state.

[0020] Pulse output module: used to extract the instruction prefetch sequence according to the dynamic buffer queue architecture, and to adjust the pulse of the instruction prefetch sequence in combination with the feed speed correction reference amount, and output the real-time interpolation control pulse for driving the servo motor.

[0021] Thirdly, the present invention provides a readable storage medium, comprising: a readable storage medium storing computer program instructions, wherein the computer program instructions are read and executed by a processor to perform the steps of a remote monitoring and communication method for CNC machine tools based on industrial Ethernet.

[0022] The beneficial effects of this invention are:

[0023] This invention constructs a delay jitter evaluation vector based on network physical time awareness. By jointly mapping it with the CNC system's reference interpolation cycle, it generates an interpolation phase difference factor and a jitter degradation index. It dynamically solves the feed speed correction reference quantity and synchronously reconstructs the machine tool's receiving buffer area. Finally, it combines the speed correction reference quantity to perform pulse adjustment on the instruction prefetch sequence and outputs real-time interpolation control pulses. This effectively solves the problem in existing technologies where random jitter in industrial Ethernet transmission delay causes a phase difference between the interpolation clock and the instruction cycle, leading to uneven servo motor pulses and resulting in physical machining marks on the workpiece surface. It achieves the suppression of the microscopic impact of network fluctuations on the transmission chain from the control source, ensuring the stability of high-speed servo motor feed, effectively eliminating irreversible machining marks caused by data cycle disorder, and improving the surface roughness index and aerodynamic service performance of aerospace precision parts.

[0024] Other features and advantages of the invention will be set forth in the following description, and will be apparent in part from the description, or may be learned by practicing the invention. The objects and other advantages of the invention can be realized and obtained by means of the structures pointed out in the description and the drawings. Attached Figure Description

[0025] Figure 1 A flowchart illustrating the remote monitoring and communication method for CNC machine tools based on industrial Ethernet of the present invention is shown.

[0026] Figure 2 This diagram illustrates the process of calculating the nonlinear velocity attenuation compensation amount according to the present invention.

[0027] Figure 3 This diagram illustrates the process of obtaining the dynamic reconfiguration addressing step size according to the present invention.

[0028] Figure 4 A schematic diagram of the modules of the CNC machine tool remote monitoring and communication system based on industrial Ethernet of the present invention is shown. Detailed Implementation

[0029] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0030] In the remote distributed manufacturing of precision aerospace parts, the control server transmits machining instructions to multiple CNC machine tools via industrial Ethernet. Precision aerospace parts typically have extremely complex free-form surface geometry and stringent dimensional tolerance requirements. During the cutting process, the CNC machine tool needs the control server to continuously send spatial coordinate data at an extremely high frequency to ensure that the cutting tool can perfectly fit the design contour of the part.

[0031] The physical environment of industrial Ethernet in actual production workshops is extremely complex. Limited by multi-level switch queuing and forwarding strategies and affected by strong electromagnetic interference caused by frequent starts and stops of large processing equipment, the physical delay of machining trajectory command data packets transmitted through the network communication medium exhibits high uncertainty. This uncertainty directly leads to microsecond-level random jitter when the machining trajectory command data packets arrive at the CNC machine tool receiver. The underlying servo control system of the CNC machine tool operates strictly at a high-precision fixed hardware clock frequency, heavily relying on a uniform and continuous data packet supply to maintain smooth physical interpolation cutting motion. The microsecond-level random jitter in the arrival time of the machining trajectory command data packets causes a significant physical phase difference between the CNC system's interpolation clock and the network command arrival cycle. When severe congestion occurs in the industrial Ethernet network of the workshop, the supply of machining trajectory command data packets may experience momentary network interruptions or arrive in a rush after a long delay, completely disrupting the originally smooth data processing rhythm of the CNC machine tool.

[0032] The mismatch between the data supply rate and the clock consumption rate causes uneven distribution of interpolation pulses when the servo motor performs high-speed feed actions. The micro-acceleration and deceleration impact generated at the end of the servo motor shaft will directly act on the cutting tool through mechanical transmission chains such as ball screws, ultimately causing irreversible physical machining marks on the surface of aerospace parts, which seriously affects the surface roughness index and aerodynamic service performance of aerospace precision parts.

[0033] This invention addresses the problem of physical machining vibration marks in the manufacturing of precision aerospace parts by providing a communication method based on underlying network physical time perception and dynamic linkage of kinematic compensation parameters with underlying cache addressing space. The aim is to suppress physical machining vibration marks caused by network fluctuations from the source of network control.

[0034] In some embodiments, such as Figure 1 As shown, this invention provides a remote monitoring and communication method for CNC machine tools based on industrial Ethernet, including:

[0035] S1. Obtain the set of physical arrival timestamps of the continuously received processing trajectory instruction data packets from the industrial Ethernet, and construct an arrival time interval sequence based on the set of physical arrival timestamps.

[0036] The system continuously monitors the physical network communication port at the bottom layer of the CNC machine tool. When a valid data frame level signal is detected, it immediately reads the current absolute physical time of the hardware clock, generates timestamp data, and reads the timestamp data in batches into the main memory. The data is then stored in the order in which the machining trajectory instruction data packets arrive, thus obtaining the set of physical arrival timestamps for the machining trajectory instruction data packets.

[0037] The actual time difference of adjacent processing trajectory instruction data packets is calculated by subtracting the timestamp value from the timestamp value that is sorted earlier in time from the timestamp value that is sorted later in time. All time difference values ​​are stored in a one-dimensional array in chronological order to construct the arrival time interval sequence. By obtaining the absolute time and performing the subtraction operation, the actual physical delay fluctuation characteristics of the industrial Ethernet network layer are captured.

[0038] For example, if recorded The arrival times of the data packets are as follows: , and .Will reduce , reduce , to obtain and The arrival time interval sequence.

[0039] S2. Extract the statistical distribution features of the arrival time interval sequence, and use the statistical distribution features to construct a network delay jitter evaluation vector that represents the current communication link congestion state.

[0040] By performing mathematical characteristic analysis such as central tendency and dispersion on the arrival time interval sequence, multi-dimensional statistical parameters are obtained. Combined with the evolution trend of statistical parameters within a specific historical time window, the multi-dimensional time feature values ​​are merged and spliced ​​into a one-dimensional array structure in the storage space to generate a network latency jitter evaluation vector that can quantitatively characterize the current industrial Ethernet congestion and jitter status.

[0041] S3. Perform joint mapping analysis between the network latency jitter evaluation vector and the preset CNC system reference interpolation period constant to generate the interpolation phase difference factor characterizing the degree of command arrival misalignment and the jitter degradation index characterizing the degree of network fluctuation.

[0042] The preset reference interpolation cycle constant of the CNC system is the absolute physical time span required for the interpolator to perform one trajectory interpolation operation, which is fixed in the underlying hardware when the CNC machine tool leaves the factory.

[0043] The network latency jitter evaluation vector is algebraically compared and proportionally converted with the inherent clock frequency constant of the CNC machine tool. The phase deviation ratio parameter affecting the position control of the servo motor is separated by specific ratio calculation. At the same time, the network disconnection probability parameter affecting the overall stability of the system is separated by threshold comparison and statistics. In this way, two independent control variables are generated, namely the interpolation phase difference factor and the jitter degradation index.

[0044] S4. Based on the interpolation phase difference factor and jitter degradation index, kinematic compensation calculation is performed to obtain the feed speed correction reference amount used to adjust the running trajectory of the servo motor.

[0045] The theoretical operating speed set in the machining program is obtained. The feedforward adjustment and nonlinear attenuation algorithm are performed in combination with the interpolation phase difference factor and the jitter degradation index to calculate the speed attenuation value that needs to be deducted to cope with microsecond-level random jitter. The speed attenuation value is subtracted from the theoretical operating speed to output the bottom line of the downgraded safe operating speed, which is the feed speed correction reference amount.

[0046] S5. Based on the feed rate correction reference amount and the jitter degradation index, the machine tool receiving buffer area is spatially reconstructed to obtain a dynamic buffer queue architecture that matches the network state.

[0047] Based on the reduced feed rate correction baseline and the jitter degradation index representing the probability of network outage, the required packet buffer byte capacity to prevent trajectory command outages is dynamically calculated. The absolute boundary addresses for physical addressing are recalculated and defined in main memory, constructing a dynamic cache queue architecture whose queuing capacity adapts to network congestion levels.

[0048] S6. Extract the instruction prefetch sequence according to the dynamic buffer queue architecture, and perform pulse adjustment on the instruction prefetch sequence in combination with the feed rate correction reference amount, and output the real-time interpolation control pulse for driving the servo motor.

[0049] The message data is read in batches from the reconstructed dynamic cache queue architecture, and the geometric execution order is rearranged according to the processing trajectory sequence number to obtain the instruction prefetch sequence.

[0050] Based on the feed rate correction reference, the single-step physical movement distance between adjacent cutting coordinate points is proportionally compressed. The compressed physical movement distance is then converted into the number of underlying electrical pulse signals, and real-time interpolation control pulses are output to the servo motor.

[0051] In some embodiments, extracting the statistical distribution features of the arrival time interval sequence includes:

[0052] S21. Calculate the arithmetic mean of all values ​​in the arrival time interval sequence.

[0053] The arithmetic mean establishes the baseline time level for data packet arrivals within the current time window, providing a reference point for subsequent measurement of jitter deviation.

[0054] For example, if the arrival time interval sequence contains , and The sum is 33, and the arithmetic mean is... .

[0055] S22. Subtract each value in the arrival time interval sequence from the arithmetic mean and square the result to obtain the square difference sequence.

[0056] Iterate through the arrival time interval sequence, subtract the arithmetic mean from each independent time difference value in the arrival time interval sequence, and calculate the physical deviation value with positive and negative signs.

[0057] Data packets may arrive early due to momentary network congestion, resulting in a negative bias, or arrive late due to queuing and forwarding by switches, resulting in a positive bias. By performing the mathematical operation of subtraction and squaring, the positive and negative cancellation effects of early and late arrival of data packets when mathematically accumulated are eliminated, and the extreme abnormal network congestion deviation value is amplified exponentially at the mathematical level.

[0058] For example, if Subtract the arithmetic mean The conclusion is ,calculate The square of the result is ,Will Subtract the arithmetic mean The conclusion is ,calculate The square of the result is ,Will Subtract the arithmetic mean The conclusion is ,calculate The square of the result is The final result includes , and The sequence of squared differences.

[0059] S23. Calculate the arithmetic mean of the squared difference sequence and output the variance eigenvalues ​​that characterize the dispersion of the arrival time interval sequence.

[0060] The variance eigenvalue is a statistical parameter that characterizes the degree of dispersion of sequence fluctuations. Through this step, the chaotic microsecond-level network physical jitter data is accurately and uniquely quantified into an evaluation parameter. The larger the value, the more uneven the arrival time of data packets.

[0061] For example, , and The sum is obtained by accumulating. ,Will Divide by the total number of data samples The variance eigenvalues ​​that characterize the degree of dispersion of the output are... .

[0062] In some embodiments, a network latency jitter evaluation vector characterizing the current communication link congestion state is constructed using statistical distribution characteristics, including:

[0063] S24. Obtain the set of variance eigenvalues ​​within the preset step size before the current time, and arrange them in chronological order as a fluctuation trend vector.

[0064] Read the historical sampling period constant that is fixed in the system configuration file. The historical sampling period constant is the preset step size.

[0065] Starting from the current absolute system clock time, the system traces back a time span equal to the preset step size in the direction of historical records, extracts all historical variance eigenvalue data recorded within the time span from the system's historical database, and obtains the variance eigenvalue set.

[0066] Read the system time tag that was attached to each data point generated in the variance feature value set, modify the address storage order of the historical variance feature value data according to the value of the time tag, and store it in a brand new 1-dimensional array to obtain the fluctuation trend vector.

[0067] The changes in the congestion status of the workshop network have both temporal continuity and physical cumulative effects. By extracting historical data and arranging it in chronological order, the stability assessment of the industrial Ethernet environment is no longer limited to a single transient state, but covers the entire cycle of network jitter evolution.

[0068] For example, to obtain the preset step size in the past Historical monitoring data was retrieved by tracing back to historical data. The variance eigenvalues ​​calculated from historical data are arranged chronologically according to time stamps to obtain a subset of... , and The fluctuation trend vector.

[0069] S25. Calculate the difference between adjacent elements in the fluctuation trend vector to obtain the fluctuation rate of change sequence.

[0070] The sudden start-up of large equipment in the workshop can cause a sharp deterioration in the network environment, which is manifested by a rapid increase in the variance eigenvalue data in a short period of time. By calculating the difference between adjacent historical elements, the gradient trend information of the intensification or weakening of network jitter can be extracted, so as to predict the possible severe deterioration of the workshop network congestion environment.

[0071] For example, reading the fluctuation trend vector, minus The result is 2, which means... minus The conclusion is ,Will and Stored in the memory address space, yielding a value. Sum of values The sequence of fluctuation rate of change.

[0072] S26. Combine the variance eigenvalues, the arithmetic mean of the fluctuation trend vector, and the terminal values ​​of the fluctuation rate of change sequence to generate a network latency jitter evaluation vector.

[0073] The arithmetic mean of the trend vector is obtained by summing the values ​​of all elements in the trend vector and then dividing by the total number of elements.

[0074] The last element in the fluctuation rate sequence is read directly from the array tail pointer. The stored element values ​​are used to derive the final value of the fluctuation rate sequence.

[0075] Allocate in main memory Includes A series of floating-point addresses In a dimensional array space, the latest calculated variance eigenvalues, the arithmetic mean of the volatility trend vector, and the terminal values ​​of the volatility rate of change sequence are respectively written into the array's dimensional array space. The, the The and the first In each indexed physical location, the numerical dimensions are combined and concatenated to obtain the network latency jitter evaluation vector.

[0076] By integrating the absolute jitter at the current moment, the average jitter level within the historical time window, and the latest jitter deterioration acceleration through address write operations, a [database / database] is constructed. A multi-dimensional and logically complete network state tensor.

[0077] For example, the numerical value of the fluctuation trend vector , , Adding them together yields , divided by The arithmetic mean of the fluctuation trend vector is obtained as follows: The final value of the fluctuation rate of change sequence is read. Obtain the latest variance eigenvalues. In the allocation Values ​​are written sequentially into the dimensional array space. , and Generate a file containing this A numerical network latency jitter evaluation vector.

[0078] In some embodiments, a joint mapping analysis is performed between the network latency jitter evaluation vector and a preset CNC system reference interpolation period constant to generate an interpolation phase difference factor characterizing the degree of command arrival misalignment and a jitter degradation index characterizing the degree of network fluctuation, including:

[0079] S31. Calculate the ratio of the variance eigenvalue of the network delay jitter evaluation vector to the square of the CNC system reference interpolation period constant to obtain the interpolation phase difference factor; count the number of values ​​in the arrival time interval sequence that are greater than twice the CNC system reference interpolation period constant to obtain the timeout count.

[0080] The ratio quantifies the control clock offset ratio at the micro level.

[0081] Multiply the reference interpolation period constant of the CNC system by The system amplifies the values ​​and uses the amplified values ​​as the threshold for judging network outages. It iterates through each time difference value in the arrival time interval sequence and compares it with the threshold. It counts the total number of values ​​greater than the threshold and obtains the timeout count. By setting a threshold of twice the constant, it accurately filters out extremely abnormal packet loss and outage risk events.

[0082] For example, if the reference interpolation period constant of the CNC system is The variance eigenvalue in the network latency jitter evaluation vector is Therefore, the interpolation phase difference factor is calculated to be 0.4, and the judgment threshold is... By traversing the arrival time interval sequence, it was found that there are 80 time difference values ​​greater than 80 in the arrival time interval sequence. The timeout count is calculated as follows. .

[0083] S32. Divide the timeout count by the total number of samples in the arrival time interval sequence to calculate the jitter degradation index.

[0084] Workshop network outages are highly sporadic and unpredictable. By dividing the absolute number of abnormal outage events by the total sample size, discrete network physical faults are smoothly transformed into continuous probability evaluation coefficients, effectively avoiding false alarms caused by CNC systems based on a single abnormal absolute value.

[0085] For example, if the timeout count is 80 and the total sample size is 1000, the jitter degradation index is calculated to be 0.08.

[0086] In some embodiments, kinematic compensation calculations are performed based on the interpolation phase difference factor and the jitter degradation index to obtain a feed rate correction reference amount for adjusting the servo motor's running trajectory, including:

[0087] S41. Obtain the command feed speed of the current interpolation cycle of the CNC system, multiply the command feed speed by the interpolation phase difference factor, and calculate the initial speed deviation value.

[0088] The trajectory interpolation planning module is accessed through the high-speed data bus inside the machine tool, and the theoretical execution speed generated by the trajectory interpolation planning module for the current control cycle is read. The theoretical execution speed is the instruction feed speed.

[0089] Multiplying the commanded feed rate by the calculated interpolation phase difference factor yields the initial speed deviation value. Network latency causes a phase difference between the arrival cycle of the spatial trajectory command and the machine tool's underlying interpolation clock, preventing the cutting tool from reaching the designated position within a specified time. Through multiplication, the abstract network latency ratio parameter is physically mapped to the absolute physical speed deviation value in the mechanical kinematic dimension, reflecting the speed tracking error that will occur if the machine tool does not intervene.

[0090] For example, if the instruction feed rate of the current interpolation cycle is The interpolation phase difference factor is The calculated initial velocity deviation value is .

[0091] S42. Obtain the maximum cutting acceleration preset by the CNC system, divide the initial speed deviation value by the maximum cutting acceleration, and calculate the theoretical buffer time.

[0092] Access the system firmware security parameter table of the servo driver and read the extreme value of the acceleration constant, which is determined by the rigidity limit of the machine tool bed casting and the peak output torque of the servo motor. The extreme value of the acceleration constant is the maximum cutting acceleration.

[0093] Divide the initial speed deviation by the maximum cutting acceleration to obtain the quotient. This quotient represents the absolute shortest time required for the machine tool's mechanical mechanism to absorb and digest the deviation under extreme working conditions, which is the theoretical buffer time. When the servo motor drives the heavy machine tool's cutting feed, an instantaneous change in physical speed will cause severe spindle vibration and damage the tool. By calculating and extracting the limit buffer time boundary through this division, it is ensured that subsequent flexible speed adjustment operations will not exceed the rigid safety constraints of the machine tool's underlying hardware.

[0094] For example, if the maximum cutting acceleration is The initial velocity deviation value is The theoretical buffer time is calculated to be... .

[0095] S43. The theoretical buffer time is made dimensionless, and a nonlinear attenuation characteristic mapping is performed by combining the command feed rate and the jitter degradation index to calculate the nonlinear speed attenuation compensation amount.

[0096] The theoretical buffer time, which contains absolute physical time units, is converted into a dimensionless coefficient by division ratio conversion. This coefficient is then combined with the command feed rate and jitter degradation index and substituted into the nonlinear mathematical model formula to perform calculations, thereby calculating the absolute speed attenuation compensation value that conforms to the machine tool's flexible start-stop characteristics.

[0097] S44. Subtract the nonlinear speed attenuation compensation amount from the commanded feed rate to calculate the output feed rate correction reference amount.

[0098] The feed rate correction reference value is the target operating speed used to safely drive the servo motor after active speed reduction and attenuation through a kinematic compensation algorithm.

[0099] For example, if the original command feed rate is... Subtract nonlinear velocity attenuation compensation amount The calculated feed rate correction reference amount is: .

[0100] In some embodiments, such as Figure 2 As shown, the theoretical buffer time is dimensionless, and a nonlinear attenuation characteristic mapping is performed by combining the command feed rate and the jitter degradation index to calculate the nonlinear speed attenuation compensation amount, including:

[0101] S431. Divide the theoretical buffer time by the CNC system reference interpolation period constant to calculate the buffer duration coefficient.

[0102] By reading the reference interpolation cycle constant of the CNC system, dividing the theoretical buffer time, which represents the shortest safe physical time, by the reference interpolation cycle constant of the CNC system, a dimensionless buffer duration coefficient is obtained, which reduces the interference of the clock physical unit scale of a specific machine tool model.

[0103] For example, dividing the theoretical buffer time of 0.02 seconds by the CNC system's reference interpolation cycle constant of 0.004 seconds yields a buffer duration coefficient of 5.

[0104] S432. Multiply the command feed rate by the jitter degradation index to obtain the degradation rate deviation, and multiply the degradation rate deviation by the buffer duration coefficient to calculate the upper limit of dynamic speed reduction.

[0105] Multiplying the original commanded feed rate by the jitter degradation index yields the expected instantaneous loss of cutting speed under conditions without any algorithmic buffering protection; this is the degradation speed deviation.

[0106] Multiply the deterioration speed deviation by the buffer duration coefficient, and use the dimensionless multiple parameter to perform numerical physical expectation amplification to obtain the limit value of the speed reduction allowed by the machine tool under the current physical state, which is the upper limit of dynamic speed reduction.

[0107] Excessive speed reduction can cause sudden changes in cutting force, leading to tool sticking. By using two rigorous multiplication calculations, the safe envelope limit of the machine tool speed reduction is defined to ensure that the speed compensation range does not exceed the dangerous bottom line that would cause cutting stall.

[0108] For example, the command feed rate With jitter degradation index Multiplying them together, we get the degradation rate deviation as follows: The rate of degradation will deviate With buffer duration coefficient Multiplying them together, we get the upper limit of the dynamic speed reduction. .

[0109] S433. Subtract the upper limit of dynamic speed reduction from the commanded feed rate to obtain the speed difference, and divide the speed difference by the commanded feed rate to calculate the output reduction compression margin.

[0110] Subtracting the upper limit of dynamic speed reduction from the commanded feed rate yields the absolute remaining speed value after deducting the extreme speed reduction protection requirement, which is the speed difference. Dividing the speed difference by the commanded feed rate converts the absolute remaining speed into a relative percentage numerical coefficient, yielding the reduction compression margin. This constructs a smooth proportional control weight factor with a constant value greater than 0 and less than 1, which drives the smooth convergence change of the subsequent secondary compensation envelope curve.

[0111] For example, the command feed rate Subtract the upper limit of dynamic speed reduction The speed difference was calculated to be... ,Will Divide by the commanded feed rate The reduction margin is calculated to be... .

[0112] S434. Substitute the upper limit of dynamic speed reduction and the reduction compression margin into the speed attenuation calculation formula to calculate and output the nonlinear speed attenuation compensation. The speed attenuation calculation formula is as follows: , This represents the nonlinear velocity attenuation compensation amount. This represents the upper limit of dynamic speed reduction. This represents the remaining margin for the reduction.

[0113] The linear deceleration strategy causes a sudden acceleration change at the moment of deceleration, which leads to flexible physical vibration of the machine tool spindle. By using specific mathematical calculations that include square convergence terms, a parabolic trajectory curve is constructed that starts smoothly at the beginning of deceleration and gradually approaches the deceleration limit at the end of deceleration. This eliminates the flexible physical vibration of the machine tool and the impact of the mechanical transmission chain caused by the instantaneous switching of linear speed.

[0114] In some embodiments, based on the feed rate correction reference amount and the jitter degradation index, the machine tool receiving buffer area is spatially reconstructed to obtain a dynamic buffer queue architecture that matches the network state, including:

[0115] S51. Obtain the preset reference buffer depth of the machine tool receiving buffer area, multiply the preset reference buffer depth by the jitter degradation index, and calculate the theoretical fluctuation compensation depth.

[0116] The machine tool receive buffer area is a contiguous storage space pre-allocated in the main physical memory of the CNC machine tool's underlying network card for temporarily storing received data packets. The maximum number of data packets statically allocated in the network kernel configuration is read; this parameter is the preset baseline buffer depth.

[0117] Multiplying the preset baseline cache depth by the jitter degradation index yields the number of additional data packet slots required to prevent disconnection under the current network congestion state, which is the theoretical fluctuation compensation depth. Instantaneous disconnection caused by network congestion will directly starve the trajectory interpolator, causing the cutting to stagnate. This calculation transforms the probability of macroscopic failure at the communication protocol layer into the actual anti-disconnection demand expansion of the underlying microscopic physical queue in terms of memory capacity.

[0118] For example, read the preset baseline cache depth as The data packets read the jitter degradation index as follows: ,Will and Multiplying these together, we obtain the theoretical fluctuation compensation depth that needs to be extended further: One data packet slot.

[0119] S52. Obtain the reference physical memory size of the machine tool receiving buffer area, and combine the theoretical fluctuation compensation depth, the reference physical memory size and the feed rate correction reference amount to perform spatial reconstruction feature mapping to obtain the dynamic reconstruction addressing step size.

[0120] Send a query command to the hardware memory controller to obtain the total number of absolute bytes of space reserved by the hardware for network communication, which is the base physical memory size.

[0121] The theoretical fluctuation compensation depth representing the queuing capacity gap, the baseline physical memory size representing the hardware physical limitation, and the feed rate correction baseline representing the machine tool's actual data digestion and consumption capacity are uniformly substituted into the system-specific addressing algebra formula to perform calculations, resulting in an integer offset value characterizing the absolute physical extension length of memory, which is the dynamic reconfiguration addressing step size.

[0122] S53. Obtain the basic physical tail address of the machine tool receiving buffer area, add the basic physical tail address to the dynamic reconstruction addressing step size to obtain the expansion physical boundary address.

[0123] Query the operating system's memory dynamic address allocation mapping table, and read the highest hexadecimal tail coordinate of the machine tool's receiving buffer area that is currently allowed to be used in terms of logical structure, which is the basic physical tail address.

[0124] The base physical tail address is added to the calculated dynamic reconfiguration addressing step size. At the physical memory level, the original memory receiving boundary is shifted towards the higher address direction to obtain the new safe endpoint hexadecimal coordinates, which is the expansion physical boundary address.

[0125] When network packets arrive in a rush, they can instantly overwhelm the default buffer. By calculating and operating on the network buffer, a coordinate barrier is established to push the network buffer backward, effectively preventing serious errors such as memory overflow loss or overwriting of important system data caused by sudden surges in data from the root.

[0126] For example, to obtain the basic physical tail address is number system The read dynamic reconstruction addressing step size is... number system ,Will and Adding them together, we get the physical boundary address of the expansion as follows: .

[0127] S54. Obtain the basic physical starting address of the machine tool receiving buffer area, construct a continuous memory block with the basic physical starting address as the starting point and the expansion physical boundary address as the ending point, and generate a dynamic cache queue architecture.

[0128] The absolute hardware starting hexadecimal coordinates of the network buffer area on the physical memory module are precisely located and read, which is the basic physical address. A high-privilege contiguous memory boundary reset mapping instruction provided by the operating system is invoked, using the basic physical address and the expansion physical boundary address as the forced start and end parameters of the instruction. This locks the new available contiguous physical space range in main memory, completing the construction of contiguous memory blocks and generating a dynamic cache queue architecture whose total capacity can adapt to the degree of external network congestion and dynamically change the physical addressing boundaries. By invoking the high-privilege instruction to reset the contiguous range of available memory, the CNC machine tool is endowed with a powerful immunity to adaptive, breathing-style memory expansion under worsening external network congestion conditions.

[0129] For example, reading the basic physical starting address is Read the physical boundary address of the expansion. The call instruction will and As start and end parameters, the physical range of the continuous interval is locked to generate a dynamic cache queue architecture.

[0130] In some embodiments, such as Figure 3 As shown, by combining the theoretical fluctuation compensation depth, the reference physical memory size, and the feed rate correction reference amount, spatial reconstruction feature mapping is performed to obtain the dynamic reconstruction addressing step size, including:

[0131] S521. Divide the theoretical fluctuation compensation depth by the baseline physical memory size to calculate the output fluctuation memory ratio.

[0132] By dividing the theoretical fluctuation compensation depth, representing the number of queued data packet demands, by the baseline physical memory size, representing the total number of bytes in the underlying hardware, the absolute data unit attribute of the physical space is stripped away, resulting in a pure numerical proportional constant that represents only the relative memory expansion intensity.

[0133] Different machine tool models have different network card memory. By calculating and generating a benchmark core adjustment ratio constant to characterize the intensity of memory resource competition, the spatial addressing reconstruction algorithm model has strong versatility for seamless cross-platform porting on CNC machine tools with different physical hardware memory configurations.

[0134] For example, the theoretical fluctuation compensation depth is read as The base physical memory size is read as ,Will Divide by The percentage of fluctuating memory was obtained as follows. .

[0135] S522. Multiply the fluctuation memory ratio by the feed rate correction baseline to calculate and output the correction rate offset.

[0136] By multiplying the fluctuating memory ratio, which represents the relative memory expansion ratio, with the feed rate correction baseline, which represents the absolute message consumption rate of the machine tool after degradation, the memory tolerance parameter and the machine tool processing speed consumption parameter are integrated into a single comprehensive dynamic variable. The slower the machine tool feed rate, the slower the network instructions are consumed, and the lower the risk of buffer accumulation. This achieves the linkage adjustment function of dynamic network resource allocation, and realizes an intelligent and reasonable closed-loop self-adjustment mechanism in which the lower the machine tool cutting data digestion speed, the safer the proportion of the underlying physical absolute memory expansion requirement.

[0137] For example, the percentage of memory read from fluctuating memory is Read the feed rate correction reference value as .Will and Multiplying them together, we get the corrected velocity offset as follows: .

[0138] S523. Obtain the preset limit feed rate of the CNC system, substitute the base physical memory size, limit feed rate, fluctuation memory ratio, and correction speed offset into the dynamic addressing calculation formula, and calculate and output the dynamic reconfiguration addressing step size. The dynamic addressing calculation formula is as follows: , Represents the addressing step size for dynamic reconstruction. Represents the baseline physical memory size. Represents the percentage of fluctuating memory. This represents the corrected velocity offset. This represents the limit feed rate.

[0139] The maximum safe movement speed physical constant limit of the machine tool, which is fixed in the configuration parameter table of the servo drive controller, is read, which is the limit feed speed.

[0140] Replace the corresponding letters in the dynamic addressing calculation formula with all extracted parameter variables, multiply the fluctuation memory ratio by the correction velocity offset, divide the product by the limit feed rate to obtain the characteristic ratio value, and add the characteristic ratio value. The overall magnification factor parameter is obtained. Multiplying the overall magnification factor parameter by the base physical memory size yields the total memory capacity occupied after the expansion. Subtracting the original base physical memory size from the total capacity occupied yields the integer increment representing the specific byte offset.

[0141] Unrestricted expansion can cause system crashes. By combining the safety ceiling constant of mechanical dynamics physical performance with the underlying network communication memory physical parameters through deep mathematical algebra, the output physical memory expansion offset byte value is guaranteed to be within the safe boundary of the machine tool's computing power that is absolutely controllable.

[0142] For example, to obtain the limit feed rate is ,calculate Multiply The conclusion is .Will Divide by The conclusion is ,add Calculate the magnification ,Will Multiply by the base physical memory size After obtaining the total capacity and subtracting the base capacity segment value, the dynamic reconfiguration addressing step size is approximately... One byte physical offset.

[0143] In some embodiments, extracting the instruction prefetch sequence according to a dynamic cache queue architecture includes:

[0144] S61. Based on the prefetch depth value determined by the dynamic cache queue architecture, sequentially read instruction data packets of the corresponding length starting from the first address of the physical memory of the machine tool receive buffer area.

[0145] The network underlying architecture control mapping table is parsed to extract the upper limit parameter of the number of instruction data packets that allow the hardware to read continuously and quickly, which is the prefetch depth value.

[0146] The underlying direct memory access hardware controller bypasses the CPU's core data bus and uses a strictly auto-incrementing physical addressing method. It directly reads the network binary data bit stream corresponding to the prefetch depth value from the physical memory starting address of the machine tool's receive buffer area, and continuously and uninterruptedly reads the bit stream in the direction of the higher memory address. The bit stream is then assembled, reassembled, and restored in the buffer area to obtain an independent instruction data packet.

[0147] High-frequency network communication generates massive data interruptions. By scheduling the direct memory access hardware controller to perform continuous hardware direct read operations, the inefficient reception response mode of triggering a central processor interrupt as soon as a single packet arrives has been changed. Hardware direct read operations greatly release the computing power of the central processor and effectively improve the overall bus throughput efficiency of CNC machine tools in centrally processing trajectory instruction data.

[0148] For example, parsing yields a prefetched depth value for extraction. Each data packet. The Direct Memory Access hardware controller continuously moves binary data bitstreams in a strictly auto-incrementing manner using physical addresses, assembling and reassembling them in memory to obtain the final data. A complete instruction data packet.

[0149] S62. Extract the sequence number field from the read instruction data packet, and reorder it according to the numerical logic of the sequence number from smallest to largest to generate the instruction prefetch sequence.

[0150] The message parsing protocol is executed to extract the strictly monotonically increasing numeric logical tag parameters that are forcibly embedded by the remote control server before sending, which are contained in the deep payload of multiple instruction data packets. These parameters are the sequence number fields.

[0151] The fast sorting algorithm embedded in the firmware is invoked. The comparison logic is executed with all extracted serial number fields as the unique comparison key. Based on the comparison results, the physical arrangement of the pointers of the read list between each data packet block in the main memory is updated to obtain the waiting queue linked list arranged according to the strict spatial processing order, which is the instruction prefetch sequence.

[0152] Interference from switch queuing and multi-hop routing can cause data packets that should be sent in order to be reversed when they arrive at the machine tool's physical network card. By extracting digital logical tag parameters and updating the order of pointers in the data block memory linked list, the disordered reading of data packets is completely corrected at the underlying data structure level. This effectively avoids the cutting tool from puncturing and colliding with the workpiece in the physical machining space due to the reversed order of instructions.

[0153] For example, the sequence number field of the five read instruction data packets has the following numeric logical labels: , , , and After calling the quicksort algorithm to perform integer comparison logic and updating the memory linked list pointer order, the result is obtained strictly according to... , , , , The physical storage stores a linked, sequential prefetch sequence of instructions.

[0154] In some embodiments, pulse adjustment is performed on the command prefetch sequence in conjunction with the feed rate correction reference amount to output real-time interpolation control pulses for driving the servo motor, including:

[0155] S63. Extract the displacement increments of adjacent trajectory points in the prefetch sequence.

[0156] Reference formula: ,in, Represents the displacement increment. and These represent the absolute geometric coordinates of two geometrically adjacent 3D spatial targets in the instruction prefetch sequence. Each movement of the CNC machine tool in each control cycle is a tiny spatial line segment. The expected physical microscopic motion length that the CNC machine tool must complete within a single low-level control cycle is obtained from the geometric data source through calculation.

[0157] For example, extract the first The starting point of a processing instruction 3D spatial coordinates , No. End point of processing instructions 3D spatial coordinates Perform the subtraction, summation, and square root operations on each axis coordinate to obtain the displacement increment. .

[0158] S64. Multiply the displacement increment by the feed rate correction reference amount to obtain the corrected single-cycle step length.

[0159] The extracted absolute linear physical geometric distance, i.e., the displacement increment, is multiplied by the feed rate correction baseline, which represents the weighting of the network security speed degradation ratio. The theoretically calculated physical movement geometric distance is then numerically compressed and scaled according to the weighting coefficients required by the anti-network jitter algorithm. This yields a parameter reflecting the absolute constraint limit of the single-cycle physical step geometric distance, which is the corrected single-cycle step length. When the network deteriorates, the machine tool cannot complete the entire process as originally planned, otherwise data interruption will occur. By directly multiplying and compressing the degraded speed macroscopic parameter with the physical microscopic geometric distance, the anti-network jitter speed protection strategy of the system's macroscopic control layer is perfectly and strictly implemented into the single-cycle physical step motion constraint of the machine tool's underlying microscopic execution layer, ensuring an absolutely smooth physical transition in the deceleration motion execution process caused by network jitter.

[0160] S65. Divide the corrected single-cycle step length by the preset pulse equivalent of the servo motor to output the number of real-time interpolation control pulses.

[0161] The constant is calculated and stored by reading the underlying electromechanical parameters. This constant is a physical constant representing the actual mechanical movement micro-distance of the machine tool corresponding to one underlying discrete electrical square wave pulse, which is obtained by electromechanical correlation calculation between the physical resolution of the servo motor photoelectric encoder hardware and the physical lead parameter of the machine tool mechanical ball screw. This is the preset pulse equivalent.

[0162] The compressed and restricted single-cycle step length is divided by the preset pulse equivalent. The division operation completely eliminates the physical units of the spatial geometric distance dimension, converting the continuous physical spatial geometric displacement into the logical number of discrete square wave pulses used to drive the motor. The integer quotient result calculated by the hardware division instruction is extracted and finally distributed to the underlying discrete electrical square wave pulse level signal of each axis servo motor driver amplifier to control the specific micro-step of the servo motor rotation. This signal is the real-time interpolation control pulse.

[0163] It generates and outputs real-time interpolation control pulses to the drive bus for physically driving the servo motor to rotate. By introducing pulse equivalents containing photoelectric encoder and lead screw parameters for physical division conversion, it ultimately completes the physical hardware and software closed loop from the upper-level software anti-network jitter delay evaluation algorithm to the lower-level high-power electrical drive execution control. It outputs the lower-level control level logic pulse signal, which effectively ensures that the electromechanical execution signal sent by the CNC machine tool always meets the damping smoothness requirements of anti-network fluctuation, thereby effectively suppressing the physical machining vibration marks on the workpiece surface.

[0164] In some embodiments, such as Figure 4 As shown, this invention provides a remote monitoring and communication system for CNC machine tools based on industrial Ethernet, comprising:

[0165] The time stamp acquisition module is used to acquire the set of physical arrival timestamps of the processing trajectory instruction data packets continuously received by the industrial Ethernet, and to construct an arrival time interval sequence based on the set of physical arrival timestamps.

[0166] Feature extraction module: used to extract the statistical distribution features of the arrival time interval sequence, and use the statistical distribution features to construct a network delay jitter evaluation vector that characterizes the current communication link congestion state.

[0167] Link evaluation module: It is used to perform joint mapping analysis between the network latency jitter evaluation vector and the preset CNC system reference interpolation period constant, and generate the interpolation phase difference factor that characterizes the degree of command arrival misalignment and the jitter degradation index that characterizes the degree of network fluctuation.

[0168] Feed compensation module: used to perform kinematic compensation calculations based on interpolation phase difference factor and jitter degradation index to obtain the feed speed correction reference amount for adjusting the running trajectory of servo motor.

[0169] Cache reconstruction module: Based on the feed rate correction baseline and jitter degradation index, it performs spatial reconstruction of the machine tool receiving buffer area to obtain a dynamic cache queue architecture that matches the network state.

[0170] Pulse output module: used to extract the instruction prefetch sequence according to the dynamic buffer queue architecture, and to adjust the pulse of the instruction prefetch sequence in combination with the feed speed correction reference amount, and output the real-time interpolation control pulse for driving the servo motor.

[0171] Thirdly, the present invention provides a readable storage medium, comprising: a readable storage medium storing computer program instructions, wherein the computer program instructions are read and executed by a processor to perform the steps of a remote monitoring and communication method for CNC machine tools based on industrial Ethernet.

[0172] It should be noted that, in this invention, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element.

[0173] Any references to memory, storage, database, or other media used in the embodiments provided in this invention may include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory may include random access memory (RAM) or external cache memory.

[0174] Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A remote monitoring and communication system for CNC machine tools based on industrial Ethernet, characterized in that, include: Time stamp acquisition module: used to acquire the set of physical arrival timestamps of continuously received processing trajectory instruction data packets from the industrial Ethernet, and to construct an arrival time interval sequence based on the set of physical arrival timestamps; Feature extraction module: used to extract the statistical distribution features of the arrival time interval sequence, and use the statistical distribution features to construct a network delay jitter evaluation vector that characterizes the current communication link congestion state; Link evaluation module: used to perform joint mapping analysis between the network latency jitter evaluation vector and the preset CNC system reference interpolation period constant, and generate interpolation phase difference factor that characterizes the degree of command arrival misalignment and jitter degradation index that characterizes the degree of network fluctuation. Feed compensation module: used to perform kinematic compensation calculations based on interpolation phase difference factor and jitter degradation index to obtain the feed speed correction reference amount for adjusting the running trajectory of servo motor; The cache reconstruction module is used to spatially reconstruct the machine tool receiving buffer area based on the feed rate correction baseline and the jitter degradation index, so as to obtain a dynamic cache queue architecture that matches the network state. Pulse output module: used to extract the instruction prefetch sequence according to the dynamic buffer queue architecture, and to adjust the pulse of the instruction prefetch sequence in combination with the feed speed correction reference amount, and output the real-time interpolation control pulse for driving the servo motor; The process of constructing a network latency jitter evaluation vector representing the current communication link congestion state using statistical distribution characteristics is as follows: Calculate the arithmetic mean of all values ​​in the arrival time interval sequence; The squared subtraction of each value in the arrival time interval sequence with the arithmetic mean yields the squared difference sequence. Calculate the arithmetic mean of the squared difference sequence, and output the variance eigenvalue that characterizes the dispersion of the arrival time interval sequence; Obtain the set of variance feature values ​​within a preset step size before the current time, and arrange them in chronological order as a fluctuation trend vector; Calculate the difference between adjacent elements in the fluctuation trend vector to obtain the fluctuation rate of change sequence; The variance eigenvalues, the arithmetic mean of the fluctuation trend vector, and the terminal values ​​of the fluctuation rate of change sequence are combined by dimension to generate a network latency jitter evaluation vector. The generation process of the interpolation phase difference factor and jitter degradation index is as follows: The interpolation phase difference factor is obtained by calculating the ratio of the variance eigenvalue of the network delay jitter evaluation vector to the square of the reference interpolation period constant of the CNC system; the timeout count is obtained by counting the number of values ​​in the arrival time interval sequence that are greater than twice the reference interpolation period constant of the CNC system. The jitter degradation index is calculated by dividing the timeout count by the total number of samples in the arrival time interval sequence. The process for obtaining the feed rate correction reference value is as follows: Obtain the command feed rate of the current interpolation cycle of the CNC system, multiply the command feed rate by the interpolation phase difference factor, and calculate the initial speed deviation value; The maximum cutting acceleration preset by the CNC system is obtained, and the initial speed deviation value is divided by the maximum cutting acceleration to calculate the theoretical buffer time. The buffer duration coefficient is calculated by dividing the theoretical buffer time by the reference interpolation cycle constant of the CNC system. Multiply the command feed rate by the jitter degradation index to obtain the degradation rate deviation, and multiply the degradation rate deviation by the buffer duration coefficient to calculate the upper limit of dynamic speed drop. Subtract the upper limit of dynamic speed reduction from the commanded feed rate to obtain the speed difference, and divide the speed difference by the commanded feed rate to calculate the output reduction compression margin. Substituting the upper limit of dynamic speed reduction and the reduction compression margin into the speed attenuation calculation formula, the nonlinear speed attenuation compensation amount is calculated and output. The speed attenuation calculation formula is as follows: , This represents the nonlinear velocity attenuation compensation amount. This represents the upper limit of dynamic speed reduction. This represents the margin for further reduction; Subtract the nonlinear speed attenuation compensation amount from the commanded feed rate to calculate and output the feed rate correction reference amount; The process of obtaining the dynamic cache queue architecture is as follows: Obtain the preset reference buffer depth of the machine tool receiving buffer area, and multiply the preset reference buffer depth by the jitter degradation index to calculate the theoretical fluctuation compensation depth; Obtain the reference physical memory size of the machine tool receiving buffer area, divide the theoretical fluctuation compensation depth by the reference physical memory size, and calculate and output the fluctuation memory ratio. Multiply the fluctuation memory ratio by the feed rate correction baseline to calculate and output the correction rate offset; Obtain the preset limit feed rate of the CNC system, substitute the reference physical memory size, limit feed rate, fluctuation memory ratio, and correction speed offset into the dynamic addressing calculation formula, and calculate and output the dynamic reconfiguration addressing step size. The dynamic addressing calculation formula is as follows: , Represents the addressing step size for dynamic reconstruction. Represents the baseline physical memory size. Represents the percentage of fluctuating memory. This represents the correction for velocity offset. Represents the limit feed rate; Obtain the basic physical tail address of the machine tool receiving buffer area, and add the basic physical tail address to the dynamic reconstruction addressing step size to obtain the expansion physical boundary address; Obtain the basic physical starting address of the machine tool receiving buffer area, construct a continuous memory block with the basic physical starting address as the starting point and the expansion physical boundary address as the ending point, and generate a dynamic cache queue architecture.

2. The system based on claim 1, characterized in that, Extracting the instruction prefetch sequence based on the dynamic cache queue architecture includes: Based on the prefetch depth value determined by the dynamic cache queue architecture, instruction data packets of the corresponding length are sequentially read starting from the first address of the physical memory of the machine tool receiving buffer area; Extract the sequence number field from the read instruction data packet and reorder it according to the numerical logic of the sequence number from smallest to largest to generate the instruction prefetch sequence.

3. The system based on claim 1, characterized in that, The command prefetch sequence is pulse-adjusted based on the feed rate correction reference, and the output is a real-time interpolation control pulse for driving the servo motor, including: Extract the displacement increments of adjacent trajectory points in the instruction prefetch sequence; Multiply the displacement increment by the feed rate correction reference amount to obtain the corrected single-cycle step length; Divide the corrected single-cycle step length by the preset pulse equivalent of the servo motor to output the number of real-time interpolation control pulses.

Images