A milling signal online processing method of processing parameter active adaptation type

Abstract

The application discloses a kind of milling signal online processing methods of processing parameter active adaptation type, including real-time reading time length t milling signal section;Determine the starting point threshold of effective signal section;By standard deviation calculation and comparison with the starting point threshold, the starting point is obtained;By standard deviation calculation and comparison determination end point threshold of effective signal section;By standard deviation calculation and comparison with the end point threshold, the end point is obtained;Judge whether the time length of effective signal section determined by starting point and end point is greater than minimum milling time length;Save the effective signal section with time length greater than minimum milling time length, return and find next effective signal section.The application realizes the active adaptation of different processing parameters by real-time automatic reading and processing of milling signal, and in addition, the application carries out online processing on the milling signal collected in real time, which not only improves the accuracy of signal processing, but also saves time cost.

Description

Technical Field

[0001] This invention belongs to the field of real-time milling signal processing technology, specifically, it relates to an online milling signal processing method with active adaptation of machining parameters. Background Technology

[0002] Machine tools are among the most important processing equipment in modern machinery manufacturing, often referred to as the "mother machine of industry." Cutting tools, acting as the "teeth" of the machine tool, are the most easily worn and failed components in the machining system, directly impacting processing quality, efficiency, energy consumption, and cost. During milling, the constant friction between the workpiece and chips inevitably leads to wear, breakage, and even fracture of the cutting tool. As tool wear intensifies, the surface quality of the machined workpiece deteriorates. Severe tool wear can result in significant waste of raw materials and even irreversible damage to the machine tool, increasing production costs. Currently, the wear condition of cutting tools is typically assessed manually based on the operator's experience and physical information generated during the cutting process. Furthermore, tool replacement is usually performed at regular intervals to prevent machining quality issues caused by tool failure.

[0003] The above judgment methods and replacement procedures ensure product quality to a certain extent, but often involve premature tool replacement. This not only increases manufacturing costs but also wastes resources, contradicting the concept of green manufacturing. On the other hand, if tool replacement is not timely, it will directly affect machining quality, leading to economic losses. To solve this problem, tool condition monitoring systems have emerged. Their core task is to monitor the tool status in real time, promptly detect and warn of abnormal tool conditions, and avoid product quality problems or even production accidents caused by tool damage. However, most machine tools process multiple products and complete multiple processes, with different machining parameters each time. Therefore, the preprocessing of milling signals and the real-time processing of signals during system use are particularly important in the development of tool condition monitoring systems, directly affecting the accuracy and sensitivity of the system.

[0004] Effective tool status information primarily exists during stable machining processes. Therefore, fine processing of milling signals—retaining only signal segments from stable milling phases and removing interference from stages such as idle feed, entry, and exit—is crucial for improving system accuracy and sensitivity. Currently, milling signal processing typically involves plotting the milling signal and manually dividing it into stable milling phases. This increases the operator's workload. Furthermore, operators often use a single threshold for judgment, which is only effective for milling signals with specific machining parameters, resulting in low accuracy. Different machining parameters require manual threshold adjustments, hindering proactive threshold adaptation. Additionally, current methods often fail to achieve online real-time signal processing; offline processing can only begin after complete signal acquisition, leading to poor real-time performance. Summary of the Invention

[0005] To overcome the above-mentioned shortcomings in the prior art, the present invention provides an active adaptive online milling signal processing method that automatically reads and processes milling signals in real time, and uses an active threshold to improve the signal processing accuracy of different machining parameters.

[0006] To achieve the above objectives, the technical solution adopted by the present invention is as follows: A method for online processing of milling signals with active adaptation of machining parameters includes the following steps: S1: Real-time reading of a milling signal segment with a duration of t; S2: Determine the threshold of the starting point of the effective signal segment by calculating the standard deviation of the signal segment read in step S1 and comparing the standard deviations of adjacent signal segments; S3: Continue reading the milling signal segment with a duration of t, and determine the quasi-starting point by calculating the standard deviation of the read signal segment and comparing it with the starting point threshold in step S2. Then wait for one cutting-in stage duration to obtain the starting point. S4: Continue reading the milling signal segment with a duration of t, and determine the effective signal segment end point threshold by calculating and comparing the standard deviation of the read signal segment. If the end point threshold cannot be determined by comparison, return to step S2. S5: After determining the threshold of the effective signal segment end point, continue to read the milling signal segment with a duration of t, and determine the quasi-end point by calculating the standard deviation of the read signal segment and comparing it with the end point threshold in step S4. Then, take one cutting stage duration forward to obtain the end point. S6: Determine whether the duration of the effective signal segment determined by the starting point obtained in step S3 and the ending point obtained in step S5 is greater than the minimum milling time; otherwise, return to step S2. S7: Save the valid signal segment in step S6 whose duration is greater than the minimum milling duration, return to step S2, and find the next valid signal segment; Wherein, the standard deviation in steps S2 to S5 is defined as std, the starting point threshold in step S2 is defined as K1, the duration of one cut-in phase in step S3 is defined as Tin, the end point threshold in step S4 is defined as K2, the duration of one cut-out phase in step S5 is defined as Tout, and the minimum milling duration in step S6 is defined as Tmin.

[0007] Further, the comparison and judgment formula in said step S3 is: std>β*K1 Wherein: std represents the standard deviation, β represents the threshold multiple, and K1 represents the starting point threshold.

[0008] Further, the comparison and judgment formula in said step S5 is: std<(1 / β)*K2 Wherein: std represents the standard deviation, β represents the threshold multiple, and K2 represents the end point threshold.

[0009] Further, said threshold multiple β is set to 5.

[0010] Further, said step S2 comprises the following steps: S201: reading a signal segment with a duration of t in real time; S202: calculating the standard deviation of the signal segment in step S201; S203: obtaining a quasi-starting point threshold through comparison and judgment with the standard deviation of the previous signal segment; S204: circularly searching subsequent signal segments with a duration of t, and performing the judgment in step S203 until the first signal segment that does not satisfy the condition is found, then proceeding to the next step; S205: obtaining that the threshold of the final starting point is the standard deviation of the last signal segment satisfying step S203.

[0011] Further, the comparison and judgment formula in said step S203 is: 0.7*std1<std<1.3std1, and α1<std<α2 Wherein: std represents the standard deviation, std1 represents the standard deviation of the previous signal segment, β represents the threshold multiple, K2 represents the end point threshold, α1 represents the lower limit value of an idle cutting signal, and α2 represents the upper limit value of an idle cutting signal.

[0012] Further, said step S4 comprises the following steps: S401: finding the starting point of the valid signal segment; S402: reading the current signal segment with a duration of t; S403: calculating the standard deviation of the signal segment with a duration of t in step S402; S404: Determine if std > α2. If yes, proceed to the next step; otherwise, return to step S2. S405: The threshold K2 of the final endpoint is the standard deviation of the signal segment that passed through step S404.

[0013] Furthermore, α1 is set to 0.1; α2 is set to 3.

[0014] Furthermore, the duration of an entry phase, Tin, is set to 4 seconds, the duration of an exit phase, Tout, is set to 4 seconds, and the minimum milling duration, Tmin, in step S6 is set to 2 seconds.

[0015] Compared with the prior art, the present invention has the following beneficial effects: (1) This invention determines the starting point threshold and the ending point threshold by judging the starting point threshold and the ending point threshold, and then further judges the starting point and the ending point by judging whether the duration of the effective signal segment determined by the starting point and the ending point is greater than the minimum milling time. The effective signal segment is obtained and saved, and then the next effective signal segment is searched. This invention automatically reads and processes the milling signal in real time, without the need for manual division by a dedicated person, reducing the burden on the operator. At the same time, the active threshold judgment realizes the active adaptation of different processing parameters, which has strong generalization and improves the processing accuracy. In addition, this invention can perform online processing on the collected signals in real time without waiting for all signals to be collected. It has strong real-time performance, which not only improves the accuracy of signal processing, but also saves time costs.

[0016] (2) In steps S3 and S5 of the present invention, by setting specific comparison and judgment formulas and setting specific threshold multiples β, the accuracy of signal processing is further improved. In addition, the present invention sets the entry stage duration Tin to 4 seconds, the exit stage duration Tout to 4 seconds, and the minimum milling duration Tmin to 2 seconds, so that the accuracy of the obtained effective signal segment is higher.

[0017] (3) The present invention obtains the quasi-starting point threshold by calculating and comparing the standard deviation of the read signal segment, and then obtains the final starting point threshold by iterative search. The whole process realizes the accurate judgment of the activity threshold. The present invention realizes the accurate judgment of the final ending point threshold by calculating the standard deviation of the read signal segment and setting the specific judgment formula std>α2. Attached Figure Description

[0018] Figure 1 This is the overall flowchart of the online milling signal processing method with active adaptation of machining parameters according to the present invention.

[0019] Figure 2The flowchart for determining the starting point threshold K1 in this invention is shown.

[0020] Figure 3 This is a flowchart for determining the end point threshold K2 in this invention.

[0021] Figure 4 This is a schematic diagram illustrating the milling signals obtained by threshold determination in this invention.

[0022] Figure 5 This is a schematic diagram of the milling signals for the idle travel stage, the entry stage, and the exit stage of the present invention. Detailed Implementation

[0023] The present invention will be further described below with reference to the accompanying drawings and embodiments. The embodiments of the present invention include, but are not limited to, the following embodiments. Example

[0024] like Figures 1 to 5 As shown, this embodiment provides an online milling signal processing method with active adaptation of machining parameters, including the following steps: S1: Real-time reading of a milling signal segment with a duration of t; S2: Determine the threshold of the starting point of the effective signal segment by calculating the standard deviation of the signal segment read in step S1 and comparing the standard deviations of adjacent signal segments; S3: Continue reading the milling signal segment with a duration of t, and determine the quasi-starting point by calculating the standard deviation of the read signal segment and comparing it with the starting point threshold in step S2. Then wait for one cutting-in stage duration to obtain the starting point. S4: Continue reading the milling signal segment with a duration of t, and determine the effective signal segment end point threshold by calculating and comparing the standard deviation of the read signal segment. If the end point threshold cannot be determined by comparison, return to step S2. S5: After determining the threshold of the effective signal segment end point, continue to read the milling signal segment with a duration of t, and determine the quasi-end point by calculating the standard deviation of the read signal segment and comparing it with the end point threshold in step S4. Then, take one cutting stage duration forward to obtain the end point. S6: Determine whether the duration of the effective signal segment determined by the starting point obtained in step S3 and the ending point obtained in step S5 is greater than the minimum milling time; otherwise, return to step S2. S7: Save the valid signal segment in step S6 whose duration is greater than the minimum milling duration, return to step S2, and search for the next valid signal segment.

[0025] In this embodiment, the standard deviation in steps S2-S5 is set to std, the starting point threshold in step S2 is set to K1, the duration of a cutting-in phase in step S3 is set to Tin, the ending point threshold in step S4 is set to K2, the duration of a cutting-out phase in step S5 is set to Tout, and the minimum milling duration in step S6 is set to Tmin.

[0026] In this embodiment, the milling signal refers to the vibration signal, current signal, force signal, and acoustic emission signal collected by sensors during the milling process. The starting point threshold and ending point threshold are determined by calculating and comparing the standard deviation of the read signal segments. Based on these thresholds, the quasi-starting point and quasi-ending point are then determined, and finally, the starting point and ending point are determined. Figure 4 As shown in the figure, this embodiment provides a schematic diagram of milling signals for determining the start and end points through threshold judgment. K1 and K2 are measured by the vertical axis, while the quasi-start point, start point, quasi-end point, and end point are measured by the horizontal axis. This embodiment performs fine processing on the milling signals, retaining only the signal segments of the stable milling stage and removing interference from the idle feed stage, the entry stage, and the exit stage. Figure 5 As shown in the figure, this embodiment provides a schematic diagram of the milling signals for the idle travel stage, the cutting-in stage, and the cutting-out stage.

[0027] In this embodiment, the comparison formula in step S3 is: std > β * K1, where std represents the standard deviation, β represents the threshold multiple, and K1 represents the starting point threshold. The comparison formula in step S5 is: std < (1 / β) * K2, where std represents the standard deviation, β represents the threshold multiple, and K2 represents the ending point threshold. Based on milling operation practice, this embodiment sets the threshold multiple β to 5, improving the accuracy of identifying valid signal segments.

[0028] In this embodiment, step S2 includes the following steps: S201: Real-time reading of signal segments with a duration of t; S202: Calculate the standard deviation of the signal segment in step S201; S203: The quasi-starting point threshold is determined by comparing it with the standard deviation of the previous signal segment; S204: Loop through the signal segments of duration t and perform the judgment in step S203 until the first signal segment does not meet the condition, then proceed to the next step; S205: The threshold for the final starting point is the standard deviation of the last signal segment that satisfies step S203.

[0029] In this embodiment, the formula for comparison and judgment in step S203 is: 0.7*std1<std<1.3*std1, and α1<std<α2, wherein std represents a standard deviation, std1 represents a standard deviation of a previous signal segment, β represents a threshold multiple, K2 represents an end point threshold, α1 represents a lower limit value of an idle cutting signal, and α2 represents an upper limit value of the idle cutting signal, as Figure 5 shown, α1 and α2 are measured by values on the ordinate.

[0030] In this embodiment, step S4 includes the following steps: S401: find a start point of the valid signal segment; S402: read a signal segment with a current duration t; S403: calculate a standard deviation of the signal segment with duration t in step S402; S404: judge whether std>α2, if yes go to the next step, if no return to step S2; S405: obtain that the threshold K2 of the final end point is the standard deviation of the signal segment that passes the judgment in step S404.

[0031] In this embodiment, based on milling practical experience, α1 is set to 0.1, α2 is set to 3, a cutting-in stage duration Tin is set to 4 seconds, a cutting-out stage duration Tout is set to 4 seconds, and a minimum milling duration Tmin in step S6 is set to 2 seconds. The setting of these values further improves the accuracy of judgment on valid signal segments. In this embodiment, the duration t is set to 0.1 seconds. The setting of a shorter duration t realizes finer reading of signal segments, and improves the judgment accuracy of a start point threshold, a quasi start point, a start point, an end point threshold, a quasi end point, and an end point.

[0032] In this embodiment, in steps S2 to S5, the standard deviation is calculated by a conventional calculation method, and the specific standard deviation calculation formula is as follows: In the formula: represents the standard deviation, N represents the total number of data, represents the n-th data, represents the average value of the data.

[0033] In this invention, the start and end thresholds of a valid signal segment are determined by calculating and comparing the standard deviation of the read signal segment with a duration of t. The quasi-start and quasi-end points are then obtained by calculating the standard deviation of the read signal segment with a duration of t and comparing it with the start and end thresholds. The quasi-start point is then waited for one entry phase duration to obtain the start point, and the quasi-end point is then taken forward for one exit phase duration to obtain the end point. Finally, it is determined whether the duration of the valid signal segment determined by the start and end points is greater than the minimum milling time. The valid signal segment is then obtained and saved, and the next valid signal segment is searched for. This invention automatically reads and processes milling signals in real time, eliminating the need for manual segmentation and reducing the burden on operators. The use of active thresholds provides strong generalization and improves processing accuracy, making it suitable for milling signals with various processing parameters. It eliminates the need for manual threshold changes, reducing personnel costs. Furthermore, this invention can process the acquired signals online in real time without waiting for all signals to be acquired, providing strong real-time performance and reducing additional operation time, thus saving time costs.

[0034] The above embodiments are merely preferred embodiments of the present invention and are not intended to limit the scope of protection of the present invention. Any changes made based on the design principles of the present invention, or any non-creative modifications made thereon, shall fall within the scope of protection of the present invention.

Claims

1. A method for online processing of milling signals with active adaptation of machining parameters, characterized in that, Comprising the following steps: S1: reading a milling signal segment with a duration of t in real time; S2: determining a starting point threshold of an effective signal segment through standard deviation calculation on the signal segment read in step S1 and standard deviation comparison and judgment between adjacent signal segments; The comparison and judgment formula in said step S3 is: std>β*K1 wherein: std represents a standard deviation, β represents a threshold multiple, and K1 represents the starting point threshold; S3: continuing to read a milling signal segment with a duration of t, determining a quasi-starting point through standard deviation calculation on the read signal segment and comparison and judgment with the starting point threshold in step S2, and then waiting for a cutting-in stage duration to obtain the starting point; S4: continuing to read a milling signal segment with a duration of t, determining an end point threshold of an effective signal segment through standard deviation calculation and comparison and judgment on the read signal segment, and returning to step S2 if a conforming end point threshold is not found; S5: after determining the end point threshold of the effective signal segment, continuing to read a milling signal segment with a duration of t, determining a quasi-end point through standard deviation calculation on the read signal segment and comparison and judgment with the end point threshold in step S4, and then taking a cutting-out stage duration forward to obtain the end point; The comparison and judgment formula in said step S5 is: std<(1 / β)*K2 wherein: std represents a standard deviation, β represents a threshold multiple, and K2 represents the end point threshold; S6: judging whether the duration of the effective signal segment determined by the starting point obtained in step S3 and the end point obtained in step S5 is greater than a minimum milling duration, and returning to step S2 if not; S7: saving the effective signal segment with a duration greater than the minimum milling duration in step S6, returning to step S2, and searching for a next effective signal segment; wherein, the standard deviation in steps S2 to S5 is set as std, the starting point threshold in step S2 is set as K1, the cutting-in stage duration in step S3 is set as Tin, the end point threshold in step S4 is set as K2, the cutting-out stage duration in step S5 is set as Tout, and the minimum milling duration in step S6 is set as Tmin.

2. The online milling signal processing method with active adaptation of machining parameters according to claim 1, characterized in that: Said threshold multiple β is set as 5.

3. The online milling signal processing method with active adaptation of machining parameters according to claim 1, characterized in that, Said step S2 comprises the following steps: S201: reading a signal segment with a duration of t in real time; S202: calculating the standard deviation of the signal segment in step S201; S203: obtaining a quasi-starting point threshold through comparison and judgment with the standard deviation of the previous signal segment; S204: circularly searching for subsequent signal segments with a duration of t, and performing the judgment in step S203 until the first signal segment that does not meet the condition is found, then proceeding to the next step; S205: obtaining that the threshold of the final starting point is the standard deviation of the last signal segment that meets the condition in step S203.

4. The online milling signal processing method with active adaptation of machining parameters according to claim 3, characterized in that, The comparison and judgment formula in said step S203 is: 0.7*std1<std<1.3std1, and α1<std<α2 wherein: std represents a standard deviation, std1 represents the standard deviation of the previous signal segment, β represents a threshold multiple, K2 represents an end point threshold, α1 represents a lower limit value of an idle cutting signal, and α2 represents an upper limit value of an idle cutting signal.

5. The online milling signal processing method with active adaptation of machining parameters according to claim 4, characterized in that, Said step S4 comprises the following steps: S401: Locate the starting point of the valid signal segment; S402: Read the signal segment with a current duration of t; S403: Calculate the standard deviation of the signal segment with duration t in step S402; S404: Determine if std > α2. If yes, proceed to the next step; otherwise, return to step S2. S405: The threshold K2 of the final endpoint is the standard deviation of the signal segment that passed through step S404.

6. The online milling signal processing method with active adaptation of machining parameters according to claim 5, characterized in that: α1 is set to 0.1; α2 is set to 3.

7. A milling signal online processing method with active adaptation of machining parameters according to any one of claims 1-6, characterized in that: Set the entry phase duration Tin to 4 seconds, the exit phase duration Tout to 4 seconds, and the minimum milling duration Tmin in step S6 to 2 seconds.

Images