【正文】 是人工神經(jīng)網(wǎng)絡(luò)的功能函數(shù)之一。人工神經(jīng)網(wǎng)絡(luò)的目標(biāo)輸出值和實(shí)際輸出值之間的差別通過學(xué)習(xí)算法悲劇算出來，并應(yīng)用方程式 1和 2調(diào)整網(wǎng)絡(luò)閾值。因此，熱電偶的分布在 200176。隱層采用雙曲線正切，對數(shù)的功能函數(shù)。在所提及的神經(jīng)網(wǎng)絡(luò)模型中，對于特殊應(yīng)用的 RP的實(shí)驗(yàn)效果最差。C ~200176。此技術(shù)在機(jī)械設(shè)備和測量領(lǐng)域具有潛在的應(yīng)用潛力。結(jié)論，本論文提出了一種基于人工神經(jīng)網(wǎng)絡(luò)的高精度溫度測量技術(shù)。在 200176。對于以上所提及的相同的網(wǎng)絡(luò)構(gòu)造，表 1顯示了應(yīng)用學(xué)習(xí)算法進(jìn)行分析的誤差。在采用不同的學(xué)習(xí)算法和和不同的網(wǎng)絡(luò)構(gòu)造后，我們發(fā)現(xiàn)最適合的網(wǎng)絡(luò)構(gòu)造是 1 X7 X 3 X 1。C 到 1000176。在我們的學(xué)習(xí)過程中應(yīng)用的神經(jīng)網(wǎng)絡(luò)算法 有 LM算法， BFGS算法， BR算法， CGF算法，和 RP算法 。訓(xùn)練完的神經(jīng)網(wǎng)絡(luò)用來測試數(shù)據(jù)但這一過程是看不見的。 3 人工神經(jīng)網(wǎng)絡(luò) 人工神 經(jīng)網(wǎng)絡(luò)是基于生物機(jī)理的人腦的模擬。此電壓作為人工神經(jīng)網(wǎng)絡(luò)的輸入值，沒有零點(diǎn)溫度補(bǔ)償?shù)臒犭娕茧妷菏侨斯ど窠?jīng)網(wǎng)絡(luò)的輸出值。相對于溫度的穩(wěn)定性是 177。具有 10V? 的輸入， AD976A的 LSB是 305A? 。這就是按照神經(jīng)元的結(jié)構(gòu)和層數(shù)和神經(jīng)元的數(shù)量通過軟件，將校正和人工神經(jīng)網(wǎng)絡(luò)仿真數(shù)據(jù)之間的均方誤差最小化。人工神經(jīng)網(wǎng)絡(luò)對 于解決復(fù)雜問題特別是非線性系統(tǒng)的仿真提供了神經(jīng)計算方法，而網(wǎng)絡(luò)本身卻是一個非線性系統(tǒng)。對于溫度，熱電偶也具有非線性輸出。C which is obtained in the calibration phase. However the performance of the final work with the training set is not an unbiased estimate of its performance on the universe of possible inputs, and a n independent test set is required to evaluate the work performance after training. Therefore, the other data set of 20 thermocouple temperatures that is uniformly distributed between 200 and 1000176。C。 stability vs. temperature: 177。C, is used in the test process. The input and output data tuples are normalized between and before training. After several trials with different learning algorithms and with different work configurations, it is found that the most suitable work configuration is 1 X 7 X 3 X 1 with the LM algorithm. This means that the number of neurons is 7 for the first hidden layer and 3 for the second hidden layer respectively. The input and output layers have the linear activation function and the hidden layers have the hyperbolic tangent sigmoid activation function. The number of epoch is 1000 for training. It is important to note that the criteria for too small and too big hidden layer neuron numbers depend on a large number of factors, like ANN type, training set characteristics and type of application. This topic is still under special attention of artificial intelligence researchers today. 4 Results and Conclusion The developed ANN models are trained and tested with the use of different learning algorithms called LM, BR, CGF, RP and BFGS to obtain better performance and faster convergence with simpler structure. Table 1 shows the errors from the plete learning algorithms used in the analysis for the same work configuration mentioned above. When the performances of the neural models are pared with each other, the best result for the training and the test are obtained from the model trained with the LM algorithm. The training and test errors (MSE, mean square error) of the work for the LM algorithm are and respectively. As it is clearly seen from Table 1, the next solution which is closer to LM is obtained from BR algorithm. Among neural models presented here, the worst results are obtained from the RP method for this particular application. It should be emphasized that the accuracy of learning algorithms in general depends on selecting appropriate learning parameters, work configurations and initializations. Figure 3 represents the percentage test error of the work trained with LM for type E thermocouple. As it is clearly seen from Figure 3, the maximum percentage error bees lower than %. The average percentage error is greater than % for temperatures between 200 and 200176。因此，傳感器的仿真和線性化技術(shù)非常必要。當(dāng)測量系統(tǒng)是非線性時包括用來訓(xùn)練的試驗(yàn)數(shù)據(jù)也是非線性時，人工神經(jīng)網(wǎng)絡(luò)是非常有用的。 2 系統(tǒng)硬件 熱電偶產(chǎn)生一和測量溫度點(diǎn)成比例的在 mV數(shù)量級的電壓值，而在零點(diǎn)的溫度值是一個常數(shù)。當(dāng)AD595作為攝氏溫度應(yīng)用時，熱電偶被忽略，微分的輸入被匯集到一起。176。在操作階段（如圖 1a），為了實(shí)現(xiàn)零點(diǎn)溫度補(bǔ)償，我們使用了從攝氏溫度計輸出的數(shù)據(jù)值。人工神經(jīng)網(wǎng)絡(luò)是前