TWI894913B - Gear fault detection system - Google Patents
Gear fault detection systemInfo
- Publication number
- TWI894913B TWI894913B TW113113952A TW113113952A TWI894913B TW I894913 B TWI894913 B TW I894913B TW 113113952 A TW113113952 A TW 113113952A TW 113113952 A TW113113952 A TW 113113952A TW I894913 B TWI894913 B TW I894913B
- Authority
- TW
- Taiwan
- Prior art keywords
- vibration
- gear
- data
- module
- time
- Prior art date
Links
Landscapes
- Testing Of Devices, Machine Parts, Or Other Structures Thereof (AREA)
- Measurement Of Mechanical Vibrations Or Ultrasonic Waves (AREA)
Abstract
Description
本申請係有關一種檢測系統，特別是指一種齒輪故障檢測系統。This application relates to a detection system, in particular to a gear fault detection system.
在習知的機械工程領域中，由於齒輪(例如正齒輪、螺旋齒輪等等)具有傳動穩定以及傳動效率高的優點，因此齒輪係為經常被使用於機械傳動的機械元件之一。In the field of mechanical engineering, gears (such as spur gears, helical gears, etc.) are one of the mechanical components often used in mechanical transmission because of their advantages of stable transmission and high transmission efficiency.
然而，隨著長時間的使用，因為機械壽命的問題，會使得齒輪開始發生磨損的狀況，進而導致齒輪的傳動效率降低及對整體機械的破壞機率提高。並且，隨著機械所要求的加工精度提高，齒輪所要求的精度也相對的提高，使得齒輪需要進行大量的檢驗及測試，才能夠投入使用，導致齒輪檢測的時間成本大幅提升。因此，如何提升齒輪的檢測效率及檢測精度便為本領域之一大課題。However, with extended use, gears can begin to wear out due to mechanical lifespan issues, reducing transmission efficiency and increasing the risk of damage to the entire machine. Furthermore, as the machining accuracy required for machinery increases, the required accuracy for gears also increases accordingly, requiring extensive inspection and testing before they can be put into use, significantly increasing the time and cost of gear testing. Therefore, improving gear inspection efficiency and accuracy has become a major issue in this field.
本申請之主要目的在於，解決舊有齒輪檢測效率及檢測精度過低的問題。The main purpose of this application is to solve the problem of low efficiency and accuracy of existing gear detection.
為達上述目的，本申請一項實施例提供一種齒輪故障檢測系統，其架設於一伺服器，一使用者透過一終端裝置訊號連接至伺服器，以使用齒輪故障檢測系統，齒輪故障檢測系統包含一齒輪偵測模組、一頻域轉換模組、一短時轉換模組、一小波轉換模組、一模型建立模組及一齒輪分析模組。齒輪偵測模組係對一齒輪對進行振動偵測，以取得一振動時域資料；頻域轉換模組耦接於齒輪偵測模組，頻域轉換模組接收振動時域資料，並透過一頻域轉換方式將振動時域資料轉換為一振動頻域資料；短時轉換模組耦接於齒輪偵測模組，短時轉換模組接收振動時域資料，並透過一短時轉換方式將振動時域資料轉換為一振動短時資料；小波轉換模組耦接於齒輪偵測模組，小波轉換模組接收振動時域資料，並透過一小波轉換方式將振動時域資料轉換為一振動小波資料；模型建立模組耦接於齒輪偵測模組、頻域轉換模組、短時轉換模組及小波轉換模組，模型建立模組接收並對振動時域資料、振動頻域資料、振動短時資料及振動小波資料進行深度學習，以取得一齒輪分析模型；齒輪分析模組耦接於齒輪偵測模組、頻域轉換模組、短時轉換模組、小波轉換模組及模型建立模組，齒輪分析模組接收並利用齒輪分析模型對振動時域資料、振動頻域資料、振動短時資料及振動小波資料進行故障分析，以取得一齒輪故障分析資料。To achieve the above-mentioned objectives, an embodiment of the present application provides a gear fault detection system, which is installed on a server. A user connects to the server via a terminal device signal to use the gear fault detection system. The gear fault detection system includes a gear detection module, a frequency domain conversion module, a short-time conversion module, a wavelet conversion module, a model building module, and a gear analysis module. The gear detection module performs vibration detection on a gear pair to obtain vibration time domain data; the frequency domain conversion module is coupled to the gear detection module, the frequency domain conversion module receives the vibration time domain data and converts the vibration time domain data into vibration frequency domain data through a frequency domain conversion method; the short time conversion module is coupled to the gear detection module, the short time conversion module receives the vibration time domain data and converts the vibration time domain data into vibration short time data through a short time conversion method; the wavelet conversion module is coupled to the gear detection module, the wavelet conversion module receives the vibration time domain data and converts the vibration time domain data into vibration wavelet data through a wavelet conversion method. data; a model building module is coupled to the gear detection module, the frequency domain conversion module, the short-time conversion module, and the wavelet conversion module. The model building module receives and performs deep learning on the vibration time domain data, the vibration frequency domain data, the vibration short-time data, and the vibration wavelet data to obtain a gear analysis model; a gear analysis module is coupled to the gear detection module, the frequency domain conversion module, the short-time conversion module, the wavelet conversion module, and the model building module. The gear analysis module receives and uses the gear analysis model to perform fault analysis on the vibration time domain data, the vibration frequency domain data, the vibration short-time data, and the vibration wavelet data to obtain gear fault analysis data.
藉此，本申請之齒輪分析模型能夠對齒輪進行多準則的檢測，預測齒輪在有磨損時會呈現何種規律，以精確地偵測齒輪的磨損狀況，並進而達到提升齒輪檢測效率及檢測精度的目的。Thus, the gear analysis model of this application can perform multi-criteria inspection on gears and predict the patterns that the gears will exhibit when they are worn, so as to accurately detect the wear condition of the gears and thereby achieve the goal of improving the gear inspection efficiency and inspection accuracy.
為便於說明本申請於上述創作內容一欄中所表示的中心思想，茲以具體實施例表達。實施例中各種不同物件係按適於列舉說明之比例，而非按實際元件的比例予以繪製，合先敘明。In order to facilitate the explanation of the central idea of this application as described in the above-mentioned creative content column, specific embodiments are presented. It should be noted that various objects in the embodiments are drawn in proportions suitable for illustration, rather than in proportion to actual components.
請參閱圖1至圖7B所示，係揭示本申請實施例之一種齒輪故障檢測系統100，其架設於一伺服器1，一使用者透過一終端裝置200訊號連接至伺服器1，以使用齒輪故障檢測系統100，齒輪故障檢測系統100包含一齒輪偵測模組10、一頻域轉換模組20、一短時轉換模組30、一小波轉換模組40、一模型建立模組50及一齒輪分析模組60。其中，齒輪故障檢測系統100係用於準確檢測齒輪的故障情況，例如齒輪的磨損程度、斷齒情況等等。1 to 7B illustrate a gear fault detection system 100 according to an embodiment of the present application. The system is deployed on a server 1. A user connects to the server 1 via a terminal device 200 to utilize the gear fault detection system 100. The gear fault detection system 100 includes a gear detection module 10, a frequency domain conversion module 20, a short-term conversion module 30, a wavelet conversion module 40, a model building module 50, and a gear analysis module 60. The gear fault detection system 100 is used to accurately detect gear fault conditions, such as gear wear and breakage.
齒輪偵測模組10，其係對一齒輪對進行振動偵測，以取得一振動時域資料。其中，所述齒輪對可由兩個正齒輪組合而成或兩個螺旋齒輪組合而成，但不以此為限，任意齒輪組合皆可為所述齒輪對；所述振動時域資料係為所述齒輪對連接於一轉動馬達，且所述齒輪對處於轉動狀態下的資料。The gear detection module 10 performs vibration detection on a gear pair to obtain vibration time-domain data. The gear pair can be composed of two spur gears or two helical gears, but this is not a limitation; any gear combination can be used. The vibration time-domain data is obtained when the gear pair is connected to a rotating motor and is in a rotating state.
頻域轉換模組20，其耦接於齒輪偵測模組10，頻域轉換模組20接收所述振動時域資料，並透過一頻域轉換方式將所述振動時域資料轉換為一振動頻域資料。其中，所述頻域轉換方式係為快速傅立葉轉換(Fast Fourier Transform, FFT)方式，所述頻域轉換方式係透過一計算式(1)將所述振動時域資料轉換為所述振動頻域資料，所述計算式(1)為： ，其中， 為所述振動頻域資料， 為所述振動時域資料， 為虛數，N為信號取樣長度。 The frequency domain conversion module 20 is coupled to the gear detection module 10. The frequency domain conversion module 20 receives the vibration time domain data and converts the vibration time domain data into vibration frequency domain data through a frequency domain conversion method. The frequency domain conversion method is a Fast Fourier Transform (FFT) method. The frequency domain conversion method converts the vibration time domain data into the vibration frequency domain data through a calculation formula (1). The calculation formula (1) is: ,in, is the vibration frequency domain data, is the vibration time domain data, is an imaginary number, and N is the signal sampling length.
短時轉換模組30，其耦接於齒輪偵測模組10，短時轉換模組30接收所述振動時域資料，並透過一短時轉換方式將所述振動時域資料轉換為一振動短時資料。其中，所述短時轉換方式係為短時傅立葉轉換(Short-time Fourier Transform, STFT)方式，所述短時轉換方式係透過一計算式(2)將所述振動時域資料轉換為所述振動短時資料，所述計算式(2)為： ，其中， 為所述振動短時資料， 為所述振動時域資料， 為窗口函數， 為虛數， 。 The short-time conversion module 30 is coupled to the gear detection module 10. The short-time conversion module 30 receives the vibration time domain data and converts the vibration time domain data into vibration short-time data through a short-time conversion method. The short-time conversion method is a short-time Fourier transform (STFT) method. The short-time conversion method converts the vibration time domain data into the vibration short-time data through a calculation formula (2). The calculation formula (2) is: ,in, is the short-term vibration data, is the vibration time domain data, is the window function, is an imaginary number. .
小波轉換模組40，其耦接於齒輪偵測模組10，小波轉換模組40接收所述振動時域資料，並透過一小波轉換方式將所述振動時域資料轉換為一振動小波資料。其中，所述小波轉換方式係為離散小波轉換(Discrete Wavelet Transform, DWT)方式，所述小波轉換方式係透過一計算式(3)將所述振動時域資料轉換為所述振動小波資料，所述計算式(3)為： ，其中， 為所述振動小波資料， 為所述振動時域資料， 為時間， 為小波函數。 The wavelet transform module 40 is coupled to the gear detection module 10. The wavelet transform module 40 receives the vibration time domain data and transforms the vibration time domain data into vibration wavelet data through a wavelet transform method. The wavelet transform method is a discrete wavelet transform (DWT) method. The wavelet transform method transforms the vibration time domain data into the vibration wavelet data through a calculation formula (3). The calculation formula (3) is: ,in, is the vibration wavelet data, is the vibration time domain data, For time, is the wavelet function.
模型建立模組50，其耦接於齒輪偵測模組10、頻域轉換模組20、短時轉換模組30及小波轉換模組40，模型建立模組50接收並對所述振動時域資料、所述振動頻域資料、所述振動短時資料及所述振動小波資料進行深度學習，以取得一齒輪分析模型51。其中，所述深度學習係透過深度神經網路(Deep Neural Networks, DNN)進行。The model building module 50 is coupled to the gear detection module 10, the frequency domain conversion module 20, the short-term conversion module 30, and the wavelet conversion module 40. The model building module 50 receives and performs deep learning on the vibration time domain data, the vibration frequency domain data, the short-term vibration data, and the vibration wavelet data to obtain a gear analysis model 51. The deep learning is performed using deep neural networks (DNNs).
更進一步的，所述深度神經網路係包含一輸入層、一隱藏層與一輸出層。所述輸入層用來接收所述振動時域資料、所述振動頻域資料、所述振動短時資料及所述振動小波資料；所述隱藏層用以將所述輸入層所接收的資料進行加權與激活，並提取資料中的特徵值；所述輸出層用以將所述隱藏層的分類後資料進行輸出，藉此建立出正確的齒輪分析模型51。Furthermore, the deep neural network includes an input layer, a hidden layer, and an output layer. The input layer is used to receive the vibration time domain data, the vibration frequency domain data, the vibration short-term data, and the vibration wavelet data; the hidden layer is used to weight and activate the data received by the input layer and extract eigenvalues from the data; and the output layer is used to output the classified data of the hidden layer, thereby establishing a correct gear analysis model 51.
齒輪分析模組60，其耦接於齒輪偵測模組10、頻域轉換模組20、短時轉換模組30、小波轉換模組40及模型建立模組50，齒輪分析模組60接收並利用齒輪分析模型51對所述振動時域資料、所述振動頻域資料、所述振動短時資料及所述振動小波資料進行故障分析，以取得一齒輪故障分析資料。所述齒輪故障分析資料係包含了各齒輪的磨損狀況以及斷齒狀況，以此讓使用者能夠明確了解是否需要進行齒輪的更換。The gear analysis module 60 is coupled to the gear detection module 10, the frequency domain conversion module 20, the short-term conversion module 30, the wavelet conversion module 40, and the model building module 50. The gear analysis module 60 receives and utilizes the gear analysis model 51 to perform fault analysis on the vibration time domain data, the vibration frequency domain data, the short-term vibration data, and the vibration wavelet data to obtain gear fault analysis data. The gear fault analysis data includes the wear and breakage status of each gear, allowing the user to clearly understand whether gear replacement is necessary.
如圖2所示，本申請更包括有一儲存模組70，其耦接於模型建立模組50及齒輪分析模組60，儲存模組70儲存所述齒輪故障分析資料，模型建立模組50能夠對所述齒輪故障分析資料進行深度學習，以精進齒輪分析模型51。As shown in FIG2 , the present application further includes a storage module 70 coupled to the model building module 50 and the gear analysis module 60 . The storage module 70 stores the gear fault analysis data, and the model building module 50 is capable of performing deep learning on the gear fault analysis data to refine the gear analysis model 51 .
請參閱圖3所示，係本申請實施例之實際實驗配置圖，包含有一齒輪對A、一馬達B及一加速規C。其中，齒輪對A係由一第一齒輪A1與一第二齒輪A2嚙合而成，第一齒輪A1及第二齒輪A2可為正齒輪或是螺旋齒輪，以此組成正齒輪對或螺旋齒輪對；馬達B係用於驅動齒輪對A進行轉動；加速規C係屬於齒輪偵測模組10的一部分，以偵測齒輪對A的所述振動時域資料。Please refer to Figure 3, which shows an actual experimental configuration diagram of an embodiment of the present application. It includes a gear pair A, a motor B, and an accelerometer C. Gear pair A is composed of a first gear A1 and a second gear A2, which can be spur gears or helical gears, forming a spur gear pair or a helical gear pair. Motor B is used to drive gear pair A to rotate. Accelerometer C is part of the gear detection module 10 and is used to detect the vibration time-domain data of gear pair A.
請參閱圖4A至圖5B所示，係為對應齒輪對A的所述振動短時資料的瀑布圖。Please refer to Figures 4A to 5B, which are waterfall diagrams of the short-term vibration data corresponding to gear pair A.
如圖4A及圖4B所示，係以齒輪對A為正齒輪對為例，圖4A係表示齒輪對A處於健康狀態，而圖4B係表示齒輪對A處於磨損狀態。其中，可看到齒輪對A處於健康狀態時，振幅區間為-20~-60，且平均振幅約為-54.1606，而齒輪對A處於磨損狀態時，振幅區間為-20~-80，且平均振幅-57.0262。相較於健康狀態的齒輪對A，處於磨損狀態的齒輪對A在振幅的呈現上，振幅區間較大且具有更多且更明顯的負向振幅，所以所述振動短時資料能夠正確的表示齒輪對A於健康狀態及磨損狀態下的振動數據差異(即當齒輪對A的振幅從-54變成-57時，代表齒輪對A出現磨損的情況)。As shown in Figures 4A and 4B, gear pair A is a spur gear pair. Figure 4A shows gear pair A in a healthy state, while Figure 4B shows gear pair A in a worn state. It can be seen that when gear pair A is in a healthy state, the amplitude ranges from -20° to -60°, with an average amplitude of approximately -54.1606°. When gear pair A is worn, the amplitude ranges from -20° to -80°, with an average amplitude of -57.0262°. Compared to the healthy gear pair A, the worn gear pair A exhibits a wider range of amplitudes and more pronounced negative amplitudes. Therefore, the short-term vibration data accurately represents the difference in vibration data between the healthy and worn states of gear pair A (i.e., when the amplitude of gear pair A changes from -54 to -57, it indicates wear).
如圖5A及圖5B所示，係以齒輪對A為螺旋齒輪對為例，圖5A係表示齒輪對A處於健康狀態，而圖5B係表示齒輪對A處於磨損狀態。其中，可看到齒輪對A處於健康狀態時，振幅區間為-20~-80，且平均振幅約為-59.7765，而齒輪對A處於磨損狀態時，振幅區間為-20~-80，且平均振幅-62.9927。相較於健康狀態的齒輪對A，處於磨損狀態的齒輪對A於時間為0.5秒~1秒時，出現最為嚴重的振幅，並且在整體的振幅呈現上，處於磨損狀態的齒輪對A具有更多且更明顯的負向振幅，所以所述振動短時資料能夠正確的表示齒輪對A於健康狀態及磨損狀態下的振動數據差異(即當齒輪對A的振幅從-59變成-62時，代表齒輪對A出現磨損的情況)。As shown in Figures 5A and 5B, using gear pair A as an example, Figure 5A shows gear pair A in a healthy state, while Figure 5B shows gear pair A in a worn state. It can be seen that when gear pair A is in a healthy state, the amplitude ranges from -20° to -80°, with an average amplitude of approximately -59.7765°. When gear pair A is worn, the amplitude ranges from -20° to -80°, with an average amplitude of -62.9927°. Compared to the healthy gear pair A, the worn gear pair A exhibits the most severe amplitude between 0.5 and 1 second. In terms of overall amplitude, the worn gear pair A exhibits more and more pronounced negative amplitudes. Therefore, the short-term vibration data accurately represents the difference in vibration data between the healthy and worn states of gear pair A (i.e., when the amplitude of gear pair A changes from -59 to -62, it indicates wear).
請參閱圖6A至圖7B所示，係為對應齒輪對A的所述振動小波資料的係數圖。Please refer to Figures 6A to 7B, which are coefficient diagrams corresponding to the vibration wavelet data of gear pair A.
如圖6A及圖6B所示，係以齒輪對A為正齒輪對為例，圖6A係表示齒輪對A處於健康狀態，而圖6B係表示齒輪對A處於磨損狀態。其中，可看到齒輪對A處於健康狀態時，平均係數為 ，而齒輪對A處於磨損狀態時，平均係數為 。因此，從所述振動小波資料可得知，當齒輪對A的平均係數從 變成 左右時，即代表齒輪對A出現磨損的情況。 As shown in Figure 6A and Figure 6B, the gear pair A is a positive gear pair. Figure 6A shows that the gear pair A is in a healthy state, while Figure 6B shows that the gear pair A is in a worn state. It can be seen that when the gear pair A is in a healthy state, the average coefficient is , and when gear pair A is in a worn state, the average coefficient is Therefore, from the vibration wavelet data, it can be seen that when the average coefficient of gear pair A changes from Become When it is around, it means that gear pair A is worn.
如圖7A及圖7B所示，係以齒輪對A為螺旋齒輪對為例，圖7A係表示齒輪對A處於健康狀態，而圖7B係表示齒輪對A處於磨損狀態。其中，可看到齒輪對A處於健康狀態時，平均係數為 ，而齒輪對A處於磨損狀態時，平均係數為 。因此，從所述振動小波資料可得知，當齒輪對A的平均係數從- 變成 左右時，即代表齒輪對A出現磨損的情況。 As shown in Figures 7A and 7B, the gear pair A is a spiral gear pair. Figure 7A shows that the gear pair A is in a healthy state, while Figure 7B shows that the gear pair A is in a worn state. It can be seen that when the gear pair A is in a healthy state, the average coefficient is , and when gear pair A is in a worn state, the average coefficient is Therefore, from the vibration wavelet data, it can be seen that when the average coefficient of gear pair A changes from - Become When it is around, it means that gear pair A is worn.
藉此，本申請之齒輪分析模型51能夠對齒輪進行多準則的檢測，預測齒輪在有磨損時會呈現何種規律，以精確地偵測齒輪的磨損狀況，並進而達到提升齒輪檢測效率及檢測精度的目的。Thus, the gear analysis model 51 of the present application can perform multi-criteria detection on the gear, predicting the pattern that the gear will exhibit when it is worn, so as to accurately detect the wear condition of the gear and thereby achieve the purpose of improving the gear detection efficiency and detection accuracy.
雖然本申請是以一個最佳實施例作說明，精於此技藝者能在不脫離本創作精神與範疇下作各種不同形式的改變。以上所舉實施例僅用以說明本創作而已，非用以限制本創作之範圍。舉凡不違本創作精神所從事的種種修改或改變，俱屬本創作申請專利範圍。Although this application illustrates a preferred embodiment, those skilled in the art will be able to make various modifications without departing from the spirit and scope of this invention. The above examples are intended only to illustrate this invention and are not intended to limit its scope. Any modifications or variations that do not violate the spirit of this invention are within the scope of this patent application.
1:伺服器 100:齒輪故障檢測系統 200:終端裝置 10:齒輪偵測模組 20:頻域轉換模組 30:短時轉換模組 40:小波轉換模組 50:模型建立模組 51:齒輪分析模型 60:齒輪分析模組 70:儲存模組 A:齒輪對 A1:第一齒輪 A2:第二齒輪 B:馬達 C:加速規 1: Server 100: Gear Fault Detection System 200: Terminal Device 10: Gear Detection Module 20: Frequency Domain Conversion Module 30: Short-Term Conversion Module 40: Wavelet Transformation Module 50: Model Building Module 51: Gear Analysis Model 60: Gear Analysis Module 70: Storage Module A: Gear Pair A1: First Gear A2: Second Gear B: Motor C: Accelerometer
[圖1]係本申請實施例之齒輪故障檢測系統之架構示意圖。 [圖2]係本申請實施例之齒輪故障檢測系統之方塊連接示意圖。 [圖3]係本申請實施例之實際實驗配置圖。 [圖4A]係本申請實施例之對應齒輪對的振動短時資料的瀑布圖，其中齒輪對為健康狀態的正齒輪對。 [圖4B]係本申請實施例之對應齒輪對的振動短時資料的瀑布圖，其中齒輪對為磨損狀態的正齒輪對。 [圖5A]係本申請實施例之對應齒輪對的振動短時資料的瀑布圖，其中齒輪對為健康狀態的螺旋齒輪對。 [圖5B]係本申請實施例之對應齒輪對的振動短時資料的瀑布圖，其中齒輪對為磨損狀態的螺旋齒輪對。 [圖6A]係本申請實施例之對應齒輪對的振動小波資料的係數圖，其中齒輪對為健康狀態的正齒輪對。 [圖6B]係本申請實施例之對應齒輪對的振動小波資料的係數圖，其中齒輪對為磨損狀態的正齒輪對。 [圖7A]係本申請實施例之對應齒輪對的振動小波資料的係數圖，其中齒輪對為健康狀態的螺旋齒輪對。 [圖7B]係本申請實施例之對應齒輪對的振動小波資料的係數圖，其中齒輪對為磨損狀態的螺旋齒輪對。 [Figure 1] is a schematic diagram of the gear fault detection system architecture of an embodiment of the present application. [Figure 2] is a schematic diagram of the block connections of the gear fault detection system of an embodiment of the present application. [Figure 3] is a diagram of the actual experimental configuration of an embodiment of the present application. [Figure 4A] is a waterfall plot of the short-term vibration data for a corresponding gear pair of an embodiment of the present application, where the gear pair is a healthy spur gear pair. [Figure 4B] is a waterfall plot of the short-term vibration data for a corresponding gear pair of an embodiment of the present application, where the gear pair is a worn spur gear pair. [Figure 5A] is a waterfall plot of the short-term vibration data for a corresponding gear pair of an embodiment of the present application, where the gear pair is a healthy helical gear pair. Figure 5B is a waterfall plot of short-term vibration data for a gear pair in accordance with an embodiment of the present application, where the gear pair is a worn helical gear pair. Figure 6A is a coefficient plot of wavelet vibration data for a gear pair in accordance with an embodiment of the present application, where the gear pair is a healthy spur gear pair. Figure 6B is a coefficient plot of wavelet vibration data for a gear pair in accordance with an embodiment of the present application, where the gear pair is a worn spur gear pair. Figure 7A is a coefficient plot of wavelet vibration data for a gear pair in accordance with an embodiment of the present application, where the gear pair is a healthy helical gear pair. Figure 7B is a coefficient diagram of the vibration wavelet data corresponding to a gear pair in an embodiment of the present application, where the gear pair is a worn helical gear pair.
100:齒輪故障檢測系統 100: Gear Fault Detection System
10:齒輪偵測模組 10: Gear Detection Module
20:頻域轉換模組 20: Frequency Domain Conversion Module
30:短時轉換模組 30: Short-term conversion module
40:小波轉換模組 40: Wavelet Transform Module
50:模型建立模組 50: Model building module
51:齒輪分析模型 51: Gear Analysis Model
60:齒輪分析模組 60: Gear Analysis Module
70:儲存模組 70: Storage Module
Claims (7)
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| TW113113952A TWI894913B (en) | 2024-04-15 | 2024-04-15 | Gear fault detection system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| TW113113952A TWI894913B (en) | 2024-04-15 | 2024-04-15 | Gear fault detection system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| TWI894913B true TWI894913B (en) | 2025-08-21 |
| TW202542494A TW202542494A (en) | 2025-11-01 |
Family
ID=97524114
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| TW113113952A TWI894913B (en) | 2024-04-15 | 2024-04-15 | Gear fault detection system |
Country Status (1)
| Country | Link |
|---|---|
| TW (1) | TWI894913B (en) |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20210148791A1 (en) * | 2019-11-14 | 2021-05-20 | Hitachi, Ltd. | Predictive maintenance for robotic arms using vibration measurements |
| TW202219796A (en) * | 2020-11-06 | 2022-05-16 | 國立陽明交通大學 | Method of obtaining vibration signal features based on machine learning model explanation |
| CN115618214A (en) * | 2022-10-14 | 2023-01-17 | 南京天洑软件有限公司 | Fault diagnosis method and device for rotating equipment |
| CN116680561A (en) * | 2023-03-29 | 2023-09-01 | 南京航空航天大学 | A Bevel Gear Fault Diagnosis Method Based on GAN-AE-LSTM under Variable Speed and Unbalanced Samples |
| TWM658331U (en) * | 2024-04-15 | 2024-07-21 | 國立勤益科技大學 | Gear Fault Detection System |
-
2024
- 2024-04-15 TW TW113113952A patent/TWI894913B/en active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20210148791A1 (en) * | 2019-11-14 | 2021-05-20 | Hitachi, Ltd. | Predictive maintenance for robotic arms using vibration measurements |
| TW202219796A (en) * | 2020-11-06 | 2022-05-16 | 國立陽明交通大學 | Method of obtaining vibration signal features based on machine learning model explanation |
| CN115618214A (en) * | 2022-10-14 | 2023-01-17 | 南京天洑软件有限公司 | Fault diagnosis method and device for rotating equipment |
| CN116680561A (en) * | 2023-03-29 | 2023-09-01 | 南京航空航天大学 | A Bevel Gear Fault Diagnosis Method Based on GAN-AE-LSTM under Variable Speed and Unbalanced Samples |
| TWM658331U (en) * | 2024-04-15 | 2024-07-21 | 國立勤益科技大學 | Gear Fault Detection System |
Also Published As
| Publication number | Publication date |
|---|---|
| TW202542494A (en) | 2025-11-01 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Tang et al. | IoT-based signal enhancement and compression method for efficient motor bearing fault diagnosis | |
| Che et al. | Partial discharge recognition based on optical fiber distributed acoustic sensing and a convolutional neural network | |
| Bo et al. | Acoustic emission detection of fatigue cracks in wind turbine blades based on blind deconvolution separation | |
| Belaid et al. | The processing of resonances excited by gear faults using continuous wavelet transform with adaptive complex Morlet wavelet and sparsity measurement | |
| CN103454113B (en) | A kind of rotating machinery health monitor method be applicable in working conditions change situation | |
| CN109710955B (en) | Fault diagnosis and health assessment method of rolling bearing based on LCD-recursive quantitative analysis | |
| Li et al. | Multipoint optimal minimum entropy deconvolution adjusted for automatic fault diagnosis of hoist bearing | |
| Hua et al. | The methodology of modified frequency band envelope kurtosis for bearing fault diagnosis | |
| CN119354540B (en) | Abnormal gearbox operation status monitoring and alarm system based on sensor analysis | |
| CN108732421A (en) | The acquisition methods and device of the instantaneous frequency of bullet train dynamic response signal | |
| TWI894913B (en) | Gear fault detection system | |
| Zhou et al. | A parameter adaptive MOMEDA method based on grasshopper optimization algorithm to extract fault features | |
| TWM658331U (en) | Gear Fault Detection System | |
| CN114878118A (en) | Transformer sound and vibration signal fusion detection method and system | |
| CN108871558A (en) | A kind of power cable operational shock health monitoring systems based on big data | |
| CN102841248B (en) | Detection method for weak signal with any frequency and range | |
| CN111259323A (en) | A method for precise location of faults of shearer rocker mechanical transmission system | |
| Zhao et al. | Fault diagnosis for gearbox based on improved empirical mode decomposition | |
| CN111220386B (en) | Bearing fault early diagnosis method based on self-adaptive frequency shift variable-scale resonance technology | |
| CN112964965A (en) | Cable fault type identification system based on frequency response | |
| CN104374576A (en) | Method for extracting fault stress waves of low-speed bearing | |
| Zhang et al. | Impulsive component extraction using shift-invariant dictionary learning and its application to gear-box bearing early fault diagnosis | |
| Du et al. | The method of quantitative trend diagnosis of rolling bearing fault based on protrugram and Lempel–Ziv | |
| CN111351645A (en) | A method for diagnosing weak fault signals of grain machinery and equipment | |
| CN116296332B (en) | Weak signal detection method based on mixed tristable stochastic resonance system |