Numerical control machine tool machining process energy sequence anomaly detection method

Abstract

The invention discloses a numerical control machine tool machining process energy sequence anomaly detection method, and belongs to the technical field of numerical control machine tool energy anomaly detection. The method specifically comprises: acquiring a standard energy sequence of a target workpiece in a numerical control machine tool machining process; searching and processing a climbing set and a downhill set in the energy sequence; dividing the energy sequence into k types of standard energy sequences based on a Gaussian hybrid clustering algorithm; obtaining a to-be-detected energy sequence, searching a climbing set and a downhill set in the energy sequence, and processing the climbing set and the downhill set; dividing the energy sequence into k types of to-be-detected energy sequences based on a Gaussian hybrid clustering algorithm; aligning the k types of to-be-detected energy sequences with the corresponding k types of standard energy sequences based on a dynamic time warping algorithm; calculating a power abnormal factor and a length abnormal factor of each type of energy sequence; and judging whether each type of energy sequence is abnormal or not. The application of the invention can reduce energy loss in industrial production and improve energy utilization efficiency.

Description

Technical Field

[0001] The invention belongs to the technical field of energy anomaly detection of numerically controlled machine tools and discloses a method for detecting energy sequence anomaly during machining of numerically controlled machine tools. Background Art

[0002] As a high-energy processing equipment, CNC milling machines have many energy components and complex system structures. Under the uncertainty of production conditions, abnormal energy consumption often occurs during the processing. In addition, due to the complexity of the system structure, CNC milling machines are prone to cascading failures. Even if the failure rate of a certain component is low, the scale effect causes the overall failure rate to increase exponentially, which in turn leads to energy abnormalities. This abnormal consumption is usually accompanied by huge energy losses and processing quality losses.

[0003] Energy full sequence anomaly (sequence anomaly for short) refers to the overall abnormality of a certain energy sequence during CNC milling. This may be caused by equipment failure, tool problems, processing parameters, operator errors or other systematic anomalies. Energy sequence anomalies in CNC milling process are divided into sequence length anomalies and sequence power anomalies. Sequence length anomalies are manifested as the length of the energy sequence being extended or shortened compared to the length of the standard energy sequence, while sequence power anomalies are manifested as the power change of the energy sequence in the power dimension being increased or decreased compared to the standard energy sequence. Therefore, it is necessary to conduct further detailed analysis of the energy sequence, taking into account the sequence anomalies.

[0004] However, existing research mostly focuses on the impact of machining parameters and tool parameters on energy and their optimization. Traditional energy monitoring and control methods also mainly alarm when the energy of a single workpiece exceeds the control limit, but cannot detect energy anomalies in the machining process in real time. In addition, common anomaly detection algorithms are easily interfered with when dealing with climbing and downhill data in the energy sequence during CNC machine tool machining, resulting in insufficient detection accuracy and stability. Summary of the invention

[0005] In view of the problems existing in the prior art, the present invention proposes a method for detecting energy sequence anomalies in a machining process of a CNC machine tool.

[0006] The present invention adopts the following technical solutions: A method for detecting abnormal energy sequence in a machining process of a numerical control machine tool, the method comprising the following steps: S1: Obtaining the standard energy sequence of the target workpiece during CNC machine tool processing .

[0007] S2: Find the standard energy sequence obtained in S1 The climbing set and the downhill set in the process are combined to obtain the processed energy sequence .

[0008] S3: Based on Gaussian mixture clustering algorithm, the energy sequence Divided into k types of standard energy sequences.

[0009] S4: Obtaining the energy sequence to be detected of the target workpiece during the CNC machine tool processing , find the energy sequence to be detected The climbing set and the downhill set in the process are combined to obtain the processed energy sequence .

[0010] S5: Based on Gaussian mixture clustering algorithm, the energy sequence Divide into k types of energy sequences to be detected.

[0011] S6: Align the k types of energy sequences to be detected with the corresponding k types of standard energy sequences based on the dynamic time warping algorithm.

[0012] S7: Calculate the power abnormality factor and length abnormality factor of each type of energy sequence; determine whether each type of energy sequence is abnormal. If the power abnormality factor or length abnormality factor of the energy sequence is greater than the set abnormal threshold, the energy sequence is determined to be abnormal.

[0013] Preferably, the standard energy sequence described in S1 refers to the energy sequence measured when a skilled operator processes the target workpiece in accordance with the standard operating instructions under ideal processing conditions, which is the standard energy sequence of the target workpiece. , ,in Indicates the time, Indicates power.

[0014] Preferably, the S2 specifically includes the following steps: S2.1: If the energy point in the energy sequence X satisfy , it is marked as the peak PV.

[0015] If the energy point in the energy sequence X satisfy , it is marked as the valley value VV.

[0016] If the energy point in the energy sequence X satisfy , it is marked as the climbing value CV.

[0017] If the energy point in the energy sequence X satisfy , it is marked as a downhill value DV.

[0018] The first energy point of the standard energy sequence is marked as a peak value, and the last energy point is marked as a valley value.

[0019] S2.2: Mark a set of single or continuous climbing values ​​as a climbing set, denoted by , the specific calculation formula is: ; ; and Represent the starting index and ending index respectively.

[0020] Mark a set of single or continuous downhill values ​​as a downhill set, denoted by , the specific calculation formula is: ; .

[0021] Collection for climbing The energy points within are assigned values, and the specific assignment formula is: .

[0022] Gather for downhill The energy points within are assigned values, and the specific assignment formula is: .

[0023] S2.3: Finally, the processed energy sequence is obtained .

[0024] Preferably, the S3 specifically includes the following steps: S3.1: Initialize the parameters of the Gaussian mixture model ; , , Represent the weight, mean, and covariance matrix of each Gaussian component in the Gaussian mixture model, respectively.

[0025] S3.2: Calculate the energy sequence based on the current Gaussian mixture model parameters Energy data in The posterior probability of belonging to class h is , the specific calculation formula is: ; Represents the Gaussian distribution of the hth component.

[0026] S3.3: Using posterior probabilities Update the weights of each Gaussian component in the Gaussian mixture model , mean and the covariance matrix , the specific calculation formula is: ; ; .

[0027] S3.4: Divided into corresponding energy sequences, assuming that each energy data The classification is , energy data The classification formula can be expressed as: .

[0028] Preferably, in S4, the energy sequence to be detected is searched The process of processing the climbing set and the downhill set in S2 is the same as that in S2, and the processed energy sequence is ; Gaussian mixture clustering algorithm transforms energy sequence The process of dividing the energy sequences to be detected into k categories is the same as that of S3.

[0029] Preferably, S6 specifically includes the following steps: S6.1: One of the standard energy sequences is , and this type of standard energy series The corresponding energy sequence to be detected is ; make ; ,in , , .

[0030] S6.2: Initialize parameters and introduce the following three constraints: Boundary: ; Monotonicity: ; Continuous type: ; in, Represents energy sequence The index of the w-th point in , Represents energy sequence The index of the w-th point in .

[0031] S6.3: Distance matrix Indicates the distance of the current energy point, , the cumulative distance matrix Indicates the distance of the current energy point The sum of the cumulative distances to the minimum neighboring energy point that can reach the energy point; use nested loops to traverse all elements of the classified energy sequence to be detected and the standard energy sequence.

[0032] S6.4: Calculate the standard energy sequence The corresponding energy sequence to be detected The similarity measure distance ; .

[0033] S6.5: The optimal alignment path set is obtained as , Connect any one of the optimal alignment paths to the energy sequence to be detected Point on The corresponding standard energy sequence Point on Path.

[0034] Preferably, S7 specifically includes the following steps: S7.1: Construct the power anomaly factor, denoted as , the specific calculation formula is: ; In the formula, Indicates Standard energy sequence on the alignment path The power value on Indicates Energy sequence to be detected on the alignment path The power value on.

[0035] S7.2: Setting the standard energy sequence The length is , energy sequence to be detected The length is , construct the energy sequence length anomaly factor, denoted as , the specific calculation formula is: .

[0036] S7.3: Setting power threshold , if the power abnormality factor Greater than the power threshold , it means that the power of the energy sequence to be detected is abnormal; on the contrary, the power of the energy sequence to be detected is normal; set the fixed length threshold of the energy sequence , if the energy sequence length anomaly factor Greater than the length threshold , it means that the length of the energy sequence to be detected is abnormal; on the contrary, it means that the length of the energy sequence to be detected is normal.

[0037] Compared with the prior art, the technical solution of the present invention has the following beneficial effects: The present invention provides a method for detecting abnormal energy sequences in the machining process of numerical control machine tools. By combining the search for climbing sets and downhill sets in the energy sequence and processing them together, Gaussian mixture clustering algorithm (GMM) and dynamic time warping algorithm (DTW), efficient and accurate detection of abnormal energy sequences in numerical control milling processes is achieved. First, the energy sequence is preprocessed to identify and smooth the peak, valley, climbing value and downhill value to reduce noise interference; secondly, the energy sequence is automatically divided according to the machining state (such as standby, spindle acceleration, no-load, cutting, spindle deceleration, etc.) through GMM to adapt to the energy change mode of different machining states; then, the DTW algorithm is used to achieve time alignment between the sequence to be detected and the standard sequence to avoid misjudgment caused by time offset; finally, by calculating the power abnormality factor and the length abnormality factor, and combining the set abnormality threshold, the power and length abnormalities of the energy sequence are comprehensively detected to ensure the accuracy and practicality of the detection results. This method significantly improves the efficiency and accuracy of energy sequence abnormality identification, can reduce energy loss in industrial production, improve energy utilization efficiency, and provide strong support for intelligent monitoring and fault diagnosis of numerical control machine tools. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] The drawings herein are incorporated into and constitute a part of the specification, illustrate embodiments consistent with the present disclosure, and together with the description are used to explain the principles of the present disclosure, and do not constitute improper limitations on the present disclosure.

[0039] Figure 1 It is a schematic diagram of the process of the present invention.

[0040] Figure 2 Energy sequence and subsequence partition diagram for CNC milling process.

[0041] Figure 3 (a) is a schematic diagram of the “climbing value” and “downhill value” of the spindle acceleration sequence; (b) is a schematic diagram of the “climbing value” and “downhill value” of the cutting sequence.

[0042] Figure 4 Schematic diagram of data assignment processing in the climbing set and the downhill set.

[0043] Figure 5 This is the energy sequence power anomaly and energy sequence length anomaly diagram.

[0044] Figure 6 This is the original energy sequence diagram of the CNC milling process in Example 1.

[0045] Figure 7 This is a result diagram of the original energy sequence of the CNC milling process in Example 1 after being processed by steps S4 and S5.

[0046] Figure 8 This is the standard energy sequence diagram of the CNC milling process in Example 1.

[0047] Fig. 9 This is the DTW alignment path diagram of the standby energy sequence in Example 1.

[0048] Fig.10 This is the DTW alignment path diagram of the spindle acceleration energy sequence in Example 1.

[0049] Fig.11 This is the DTW alignment path diagram of the no-load energy sequence in Example 1.

[0050] Fig.12 This is the DTW alignment path diagram of the cutting energy sequence in Example 1.

[0051] Fig.13 This is the DTW alignment path diagram of the spindle deceleration energy sequence in Example 1. DETAILED DESCRIPTION

[0052] The specific implementation of the present invention is further described below in conjunction with the accompanying drawings and specific examples.

[0053] Combination Figures 1 to 5 , a method for detecting abnormal energy sequence in a CNC machine tool machining process, characterized in that the method comprises the following steps: S1: Obtaining the standard energy sequence of the target workpiece during CNC machine tool processing .

[0054] The standard energy sequence is the energy sequence measured when a skilled operator processes the target workpiece according to the standard operating instructions under ideal processing conditions. , where each energy point ,in Indicates the time, Indicates power.

[0055] like Figure 2 As shown in the figure, from the perspective of energy flow and consumption in the CNC milling process, taking the CNC milling plane as an example, the CNC milling process usually includes shutdown, startup, standby, spindle acceleration, no-load, cutting and spindle deceleration, etc. Among them, there is no energy consumption during shutdown and although the transient power of startup is large, the number of startups is small and the startup time period is short. Therefore, the shutdown and startup processes are not considered temporarily. Therefore, the CNC machine tool processing process can be divided into standby sequence, spindle acceleration sequence, no-load sequence, cutting sequence and spindle deceleration sequence.

[0056] An energy sequence may have one or more energy subsequences. For example, a CNC milling plane process has two standby energy subsequences in the standby energy sequence, four no-load energy subsequences in the no-load energy sequence, three cutting energy subsequences in the cutting energy sequence, and the spindle acceleration and deceleration energy sequences each contain an energy subsequence, that is, an energy subsequence is an energy sequence. In addition, considering the uncertain abnormal energy in the CNC milling process, the abnormal energy fragment composed of ≥2 consecutive abnormal energy points is also regarded as an energy subsequence. It is worth noting that in the actual processing process, the energy sequence of the CNC milling process needs to be divided according to the actual processing state.

[0057] (1) Standby sequence: In milling, the standby sequence refers to the period when the machine is powered but not actively involved in the machining process. In the standby sequence, the machine is ready but not actively involved in cutting. Power consumption is relatively low, but not negligible. Power may be used to maintain the standby functions of the machine, including keeping the control system active. (2) Spindle acceleration sequence: The spindle acceleration sequence involves the machine moving at a high speed to reach the desired position before the actual milling operation begins. When the milling process begins, the spindle begins to accelerate, and additional power is required to overcome inertia and friction, and power consumption increases. Energy is used to accelerate the spindle to the required speed. The power trend during this phase is upward until the spindle reaches the required rotational speed. (3) No-load sequence: The no-load sequence occurs when the milling machine is in operation but not involved in the cutting operation. Once the spindle reaches its operating speed and is not cutting (no-load state), power consumption tends to stabilize at a relatively constant level. The machine is running, but the cutting tool is not involved in material removal. Power is used to maintain spindle speed and other machine functions. (4) Cutting sequence: The cutting sequence is the core of the milling process and is the stage in which material is removed from the workpiece by rotating the cutting tool. The cutting state involves actively milling or cutting the workpiece. The power consumption in this stage is generally higher than that in the no-load state. The power trend will fluctuate depending on factors such as the cutting material, cutting tool specifications, and cutting depth and speed. Higher material removal rates or harder materials may result in increased power consumption. (5) Spindle deceleration sequence: The spindle deceleration sequence involves withdrawing the cutting tool at high speed after the machining operation is completed. As the milling process ends, the spindle begins to decelerate and the power consumption decreases. Energy is used to gradually stop the spindle. The power trend in this state is downward until the spindle comes to a complete stop.

[0058] S2: Find the standard energy sequence obtained in S1 The climbing set and the downhill set in the process are combined to obtain the processed energy sequence .

[0059] In the process of CNC milling, there are a lot of special change trends in the time series. Figure 3(a), due to transient acceleration, the power of the spindle acceleration sequence shows a trend of first increasing and then decreasing. This trend will inevitably lead to the appearance of "climbing values" and "downhill values" in the original time series under the spindle acceleration state; in addition, similar data features also exist in the spindle deceleration sequence and cutting sequence ( Figure 3 (b)). It should be noted that the reason for this type of data in the cutting sequence is different from that in the spindle acceleration / deceleration sequence. The spindle acceleration / deceleration sequence is caused by the transient change of the spindle speed, while the cutting sequence is caused by the incomplete contact between the tool and the workpiece. This type of data will seriously affect the judgment of the processing sequence. Therefore, without affecting the amount of processing sequence data, eliminating the "climbing value" and "downhill value" in the original time series is helpful for online judgment of the processing sequence.

[0060] The rainflow counting method was first proposed by TN Frosch and NEM Morrow in 1968. It is mainly used for fatigue analysis to evaluate the fatigue damage of materials or structures under cyclic loads. This method decomposes the load history into a series of non-overlapping cyclic loads, and then counts and analyzes these cyclic loads. The basic principle is to identify cyclic loads based on the peak and valley values ​​of the load history and count these cyclic loads. Based on the principle of the rainflow counting method and the energy characteristics of the CNC milling process, the present invention proposes to find the climbing set and the downhill set in the energy sequence and process them, which mainly includes the following steps: 1) Labeling data types: Mark each data in the original time series as a climbing value, downhill value, peak value or valley value. 2) Finding climbing sets and downhill sets: Traverse the data type of each data in the original time series, and form a data set with continuous data of the same data. 3) Assignment processing: Assign a new value to the data in the climbing set and the downhill set. Specifically includes the following steps: S2.1: If the energy point in the energy sequence X satisfy , it is marked as the peak PV.

[0061] If the energy point in the energy sequence X satisfy , it is marked as the valley value VV.

[0062] If the energy point in the energy sequence X satisfy , it is marked as the climbing value CV.

[0063] If the energy point in the energy sequence X satisfy , it is marked as a downhill value DV.

[0064] The first energy point of the standard energy sequence is marked as a peak value, and the last energy point is marked as a valley value.

[0065] S2.2: Mark a set of single or continuous climbing values ​​as a climbing set, denoted by , the specific calculation formula is: ; .

[0066] Mark a set of single or continuous downhill values ​​as a downhill set, denoted by , the specific calculation formula is: ; .

[0067] and Represent the starting index and the ending index respectively. It is worth noting that when = When , it means that the "slope value" is discontinuous, that is, an independent "slope value" constitutes a "slope"; when When , it means that the "slope value" is continuous and uninterrupted, that is, multiple consecutive "slope values" constitute a "slope". According to the definition of "slope", there is no "slope" that has both "climbing value" and "downhill value".

[0068] Collection for climbing The energy points within are assigned values, and the specific assignment formula is: .

[0069] Gather for downhill The energy points within are assigned values, and the specific assignment formula is: .

[0070] The diagram of assignment processing is as follows Figure 4 .

[0071] S2.3: Finally, the processed energy sequence is obtained .

[0072] During CNC milling, the machine tool's spindle acceleration, spindle deceleration, and cutting power data show climbing or downhill characteristics. In order to process this type of raw time series energy data, the peak and valley time series is obtained by replacing the climbing and downhill values. The purpose of this operation is to eliminate the interference of these data on the discrimination of the processing sequence and improve the accuracy of subsequent abnormality detection.

[0073] S3: Based on Gaussian mixture clustering algorithm, the energy sequence Divided into k types of standard energy sequences.

[0074] The k value here needs to be specified according to the actual machining process. In this embodiment, the actual CNC milling process includes five machining sequences, namely, a standby sequence, a spindle acceleration sequence, an idle sequence, a cutting sequence, and a spindle deceleration sequence; therefore, k is 5.

[0075] S3 specifically includes the following steps: S3.1: Initialize the parameters of the Gaussian mixture model .

[0076] , , Represent the weight, mean, and covariance matrix of each Gaussian component in the Gaussian mixture model, respectively.

[0077] S3.2: Calculate the energy sequence based on the current Gaussian mixture model parameters Energy data in The posterior probability of belonging to class h is , the specific calculation formula is: .

[0078] Represents the Gaussian distribution of the hth component, that is, the probability density function.

[0079] S3.3: Using posterior probabilities Update the weights of each Gaussian component in the Gaussian mixture model , mean and the covariance matrix , the specific calculation formula is: .

[0080] .

[0081] .

[0082] S3.4: Divided into corresponding energy sequences, assuming that each energy data The classification is , energy data The classification formula can be expressed as: .

[0083] S4: Obtaining the energy sequence to be detected of the target workpiece during the CNC machine tool processing , find the energy sequence to be detected The climbing set and the downhill set in the process are combined to obtain the processed energy sequence .

[0084] Find the energy sequence to be detected in S4 The process of processing the climbing set and the downhill set in S2 is the same as that in S2, and the processed energy sequence is .

[0085] S5: Based on Gaussian mixture clustering algorithm, the energy sequence The energy sequences are divided into five categories to be detected, and the Gaussian mixture clustering algorithm is used to classify the energy sequences The process of dividing the energy sequences to be detected into k categories is the same as that of S3.

[0086] S6: Align the k types of energy sequences to be detected with the corresponding k types of standard energy sequences based on the dynamic time warping algorithm.

[0087] Due to external factors such as changes in the operating environment, equipment differences or human factors, or internal factors such as processing parameter errors, each processing sequence in the actual CNC milling process may have a slight time offset, distortion or compression with its corresponding standard processing sequence. In order to more accurately determine whether the processing sequence is abnormal, it is necessary to align the measured processing sequence with the standard processing sequence to ensure that they match better. To solve this alignment problem, the present invention uses the dynamic time warping algorithm (DTW) method. DTW is a flexible time series alignment method that finds the optimal alignment path by performing nonlinear stretching or compression on the time axis, thereby minimizing the distance between the two sequences.

[0088] S6 specifically includes the following steps: S6.1: One of the standard energy sequences is , and this type of standard energy series The corresponding energy sequence to be detected is ; make ; ,in , , .

[0089] S6.2: Initialize parameters and introduce the following three constraints.

[0090] Boundary: ; Indicates that the beginning and end of the two sequences must match, and the order of each part must match.

[0091] Monotonicity: ; Indicates that the path must be monotonically increasing over time.

[0092] Continuous type: ; This constraint means that in the many-to-one and one-to-many matching process, only the surrounding time step can be matched, that is, it is impossible to match across a certain point, and it can only be aligned with the adjacent points.

[0093] in, Represents energy sequence The index of the w-th point in , Represents energy sequence The index of the w-th point in .

[0094] S6.3: Distance matrix Indicates the distance of the current energy point, , the cumulative distance matrix Indicates the distance of the current energy point The sum of the cumulative distances to the minimum neighboring energy point that can reach the energy point; use nested loops to traverse all elements of the classified energy sequence to be detected and the standard energy sequence.

[0095] S6.4: Calculate the standard energy sequence The corresponding energy sequence to be detected The similarity measure distance ; .

[0096] S6.5: The optimal alignment path set is obtained as , Connect any one of the optimal alignment paths to the energy sequence to be detected Point on The corresponding standard energy sequence Point on Path.

[0097] S7: Calculate the power abnormality factor and length abnormality factor of each type of energy sequence; determine whether each type of energy sequence is abnormal. If the power abnormality factor or length abnormality factor of the energy sequence is greater than the set abnormal threshold, the energy sequence is determined to be abnormal.

[0098] Sequence abnormality refers to the energy abnormality of a certain processing sequence as a whole during CNC milling, such as Figure 5As shown. This may indicate that the energy consumption of the entire processing sequence is abnormal as a whole, which may be caused by equipment failure, tool problems, processing parameters, operator errors or other systematic abnormalities. It is worth noting that in the present invention, the sequence abnormality of the CNC milling process is divided into sequence length abnormality and sequence power abnormality. The sequence length abnormality is manifested as the length of the processing sequence being extended or shortened compared to the length of the standard processing sequence, while the sequence power abnormality is manifested as the power change of the processing sequence in the power dimension compared to the standard processing sequence. Therefore, there may still be sequence abnormalities even if there is no point abnormality or subsequence abnormality on the power curve. For example, (a) when the cutting depth is greater than the given standard cutting depth, the power of the cutting sequence becomes larger than the standard cutting power; (b) when the standby time is greater than the standard standby time, the length of the standby sequence is greater than the length of the standard standby sequence. It is not difficult to understand that abnormality (a) is a sequence power abnormality; and abnormality (b) is a sequence length abnormality.

[0099] S7 specifically includes the following steps: S7.1: Construct the power anomaly factor, denoted as , the specific calculation formula is: .

[0100] In the formula, Indicates Standard energy sequence on the alignment path The power value on Indicates Energy sequence to be detected on the alignment path The power value on.

[0101] S7.2: Setting the standard energy sequence The length is , energy sequence to be detected The length is , construct the energy sequence length anomaly factor, denoted as , the specific calculation formula is: .

[0102] S7.3: Setting power threshold , if the power abnormality factor Greater than the power threshold , it means that the power of the energy sequence to be detected is abnormal; on the contrary, the power of the energy sequence to be detected is normal; set the fixed length threshold of the energy sequence , if the energy sequence length anomaly factor Greater than the length threshold , it means that the length of the energy sequence to be detected is abnormal; on the contrary, it means that the length of the energy sequence to be detected is normal.

[0103] Example 1 Combination Figures 6 to 13 During a CNC milling process, the operator mistakenly set the spindle speed to 6000r / min, resulting in the original time series of the CNC milling process as follows Figure 6 shown. Figure 7 The results after processing in steps S4 and S5 are shown, which clearly shows that the CNC milling process is divided into a standby sequence, an idle sequence, a cutting sequence, a spindle acceleration sequence, and a spindle deceleration sequence.

[0104] The standard cutting depth is 5mm, the standard feed speed is 600mm / s, the standard spindle speed is 8000r / min, and the standard cutting depth is 1mm. The standard processing sequence is as follows: Figure 8 shown.

[0105] The length abnormality threshold of the standby energy sequence, spindle acceleration energy sequence, no-load energy sequence, cutting energy sequence and spindle deceleration energy sequence is set to 0.3; the power abnormality threshold of the standby energy sequence, no-load energy sequence and cutting energy sequence is set to 30; the power abnormality threshold of the spindle acceleration energy sequence and the spindle deceleration energy sequence is set to 500.

[0106] Figure 9-13 In the figure, the blue dots represent the power points on the sequence to be tested, and the red dots represent the power points of the standard processing sequence.

[0107] In this embodiment, if Fig. 9 As shown, the DTW alignment path of the standby energy sequence is displayed. The detection results show that the length of the standard standby energy sequence is 264, while the length of the instance standby energy sequence is 279, and the length anomaly factor is 0.057, which is less than the length anomaly threshold, indicating that the energy sequence length is normal; the power anomaly factor of the instance standby energy sequence is 0.519, which is less than the sequence power anomaly threshold, indicating that the energy sequence power is normal.

[0108] In this embodiment, if Fig.10 As shown in the figure, the DTW alignment path of the spindle acceleration energy sequence is shown. The detection results show that the length of the standard spindle acceleration energy sequence is 17, while the length of the case spindle acceleration energy sequence is 21, and the length anomaly factor is 0.235, which is less than the length anomaly threshold, indicating that the length of the energy sequence is normal; the power anomaly factor of the energy sequence is 262.261, which is less than the sequence power anomaly threshold, indicating that the power of the energy sequence is normal.

[0109] In this embodiment, if Fig.11As shown in the figure, the DTW alignment path of the no-load energy sequence is shown. The detection results show that the length of the standard no-load energy sequence is 182, while the length of the case no-load energy sequence is 178, and the length anomaly factor is 0.022, which is less than the length anomaly threshold, indicating that the energy sequence length is normal; the energy sequence power anomaly factor is 59.052, which is greater than the sequence power anomaly threshold, indicating that the energy sequence power is abnormal.

[0110] In this embodiment, if Fig.12 As shown in the figure, the DTW alignment path of the cutting energy sequence is shown. The detection results show that the length of the standard cutting energy sequence is 334, while the length of the case cutting energy sequence is 339, and the length anomaly factor is 0.015, which is less than the length anomaly threshold, indicating that the length of the energy sequence is normal; the power anomaly factor of the case cutting energy sequence is 34.659, which is greater than the sequence power anomaly threshold, indicating that the power of the energy sequence is abnormal.

[0111] In this embodiment, if Fig.13 As shown in the figure, the DTW alignment path of the spindle deceleration energy sequence is shown. The detection results show that the length of the standard spindle deceleration energy sequence is 14, while the length of the case spindle deceleration energy sequence is 16, and the length abnormality factor is 0.143, which is less than the length abnormality threshold, indicating that the length of the energy sequence is normal; the power abnormality factor of the case cutting energy sequence is 248.412, which is less than the sequence power abnormality threshold, indicating that the power of the energy sequence is normal.

[0112] On the whole, in this embodiment, the abnormality detection results of the energy sequence of the CNC milling process are as follows: the power and length of the standby energy sequence are normal; the power and length of the spindle acceleration energy sequence are normal; the no-load energy sequence length is normal, but the power is abnormal; the cutting energy length is normal, but the power is abnormal; the power and length of the spindle deceleration energy sequence are normal. The above experimental results are consistent with the actual results, verifying the reliability of the detection method. The energy sequence abnormality detection method for the CNC machining process proposed in the present invention has significant effects, and can accurately divide the energy sequence of the CNC milling process into several types of energy sequences according to the machining state, and at the same time realize the alignment of the standard energy sequence and the sequence to be detected, thereby realizing the abnormality detection of each energy sequence separately, which helps to improve the efficiency and accuracy of energy sequence abnormality identification.

[0113] Of course, the above description is not a limitation of the present invention, and the present invention is not limited to the above examples. Changes, modifications, additions or substitutions made by technicians in this technical field within the essential scope of the present invention should also fall within the protection scope of the present invention.

Claims

1. A method for detecting abnormal energy sequence in a CNC machine tool machining process, characterized in that: The method comprises the following steps: S1: Obtaining the standard energy sequence of the target workpiece during CNC machine tool processing ; S2: Find the standard energy sequence obtained in S1 The climbing set and the downhill set in the process are combined to obtain the processed energy sequence ; S3: Based on Gaussian mixture clustering algorithm, the energy sequence Divided into k types of standard energy sequences; S4: Obtaining the energy sequence to be detected of the target workpiece during the CNC machine tool processing , find the energy sequence to be detected The climbing set and the downhill set in the process are combined to obtain the processed energy sequence ; S5: Based on Gaussian mixture clustering algorithm, the energy sequence Divide into k types of energy sequences to be detected; S6: aligning k types of energy sequences to be detected with corresponding k types of standard energy sequences based on a dynamic time warping algorithm; S7: Calculate the power abnormality factor and length abnormality factor of each type of energy sequence; determine whether each type of energy sequence is abnormal. If the power abnormality factor or length abnormality factor of the energy sequence is greater than the set abnormal threshold, the energy sequence is determined to be abnormal.

2. The method for detecting abnormal energy sequence in a CNC machine tool machining process according to claim 1, characterized in that: The standard energy sequence described in S1 refers to the energy sequence measured when a skilled operator processes the target workpiece in accordance with the standard operating instructions under ideal processing conditions, which is the standard energy sequence of the target workpiece. , ,in Indicates the time, Indicates power.

3. The method for detecting abnormal energy sequence in a CNC machine tool machining process according to claim 2, characterized in that: The S2 specifically includes the following steps: S2.1: If the energy point in the energy sequence X satisfy , then mark it as the peak PV; If the energy point in the energy sequence X satisfy , then mark it as the valley value VV; If the energy point in the energy sequence X satisfy , then mark it as the climbing value CV; If the energy point in the energy sequence X satisfy , then mark it as downhill value DV; Among them, the first energy point of the standard energy sequence is marked as the peak value, and the last energy point is marked as the valley value; S2.2: Mark a set of single or continuous climbing values ​​as a climbing set, denoted by , the specific calculation formula is: ； ； and Represent the starting index and the ending index respectively; Mark a set of single or continuous downhill values ​​as a downhill set, denoted by , the specific calculation formula is: ； ； Collection for climbing The energy points in the array are assigned values. The specific assignment formula is: ； Gather for downhill The energy points in the array are assigned values. The specific assignment formula is: ； S2.3: Finally get the processed energy sequence .

4. The method for detecting abnormal energy sequence in a CNC machine tool machining process according to claim 3, characterized in that: The S3 specifically includes the following steps: S3.1: Initialize the parameters of the Gaussian mixture model ; , , Respectively represent the weight, mean and covariance matrix of each Gaussian component in the Gaussian mixture model; S3.2: Calculate the energy sequence based on the current Gaussian mixture model parameters Energy data in The posterior probability of belonging to class h is , the specific calculation formula is: ； represents the Gaussian distribution of the hth component; S3.3: Using posterior probabilities Update the weights of each Gaussian component in the Gaussian mixture model , mean and the covariance matrix , the specific calculation formula is: ； ； ； S3.4: Divided into corresponding energy sequences, assuming that each energy data The classification is , energy data The classification formula can be expressed as: 。 5. The method for detecting abnormal energy sequence in a CNC machine tool machining process according to claim 2, characterized in that: Find the energy sequence to be detected in S4 The process of processing the climbing set and the downhill set in S2 is the same as that in S2, and the processed energy sequence is ; Gaussian mixture clustering algorithm transforms energy sequence The process of dividing the energy sequences to be detected into k categories is the same as that of S3.

6. A method for detecting abnormal energy sequence in a CNC machine tool machining process according to claim 5, characterized in that: S6 specifically includes the following steps: S6.1: One of the standard energy sequences is , and this type of standard energy series The corresponding energy sequence to be detected is ; make ; ,in , , ; S6.2: Initialize parameters and introduce the following three constraints: Boundary: ; Monotonicity: ; Continuous type: ; in, Represents energy sequence The index of the w-th point in , Represents energy sequence The index of the w-th point in ; S6.3: Distance matrix Indicates the distance of the current energy point, , the cumulative distance matrix Indicates the distance of the current energy point The sum of the cumulative distances to the minimum neighboring energy point that can reach the energy point; using nested loops to traverse all elements of the energy sequence to be detected and the standard energy sequence after classification; S6.4: Calculate the standard energy sequence The corresponding energy sequence to be detected The similarity measure distance ; ； S6.5: The optimal alignment path set is obtained as , Connect any one of the optimal alignment paths to the energy sequence to be detected Point on The corresponding standard energy sequence Point on Path.

7. A method for detecting abnormal energy sequence in a CNC machine tool machining process according to claim 6, characterized in that: S7 specifically includes the following steps: S7.1: Construct the power anomaly factor, denoted as , the specific calculation formula is: ； In the formula, Indicates Standard energy sequence on the alignment path The power value on Indicates Energy sequence to be detected on the alignment path The power value on S7.2: Setting the standard energy sequence The length is , energy sequence to be detected The length is , construct the energy sequence length anomaly factor, denoted as , the specific calculation formula is: ； S7.3: Setting power threshold , if the power abnormality factor Greater than the power threshold , it means that the power of the energy sequence to be detected is abnormal; on the contrary, the power of the energy sequence to be detected is normal; set the fixed length threshold of the energy sequence , if the energy sequence length anomaly factor Greater than the length threshold , it means that the length of the energy sequence to be detected is abnormal; on the contrary, it means that the length of the energy sequence to be detected is normal.

Images