Laser-induced breakdown spectroscopy (LIBS) coupled with chemometrics is an efficient method for rock identification and classification, which has considerable potential in planetary geology. A great challenge facing the LIBS community is the difficulty to accurately discriminate rocks with close chemical compositions. A convolutional neural network (CNN) model has been designed in this study to identify twelve types of rock, among which some rocks have similar compositions. Both the training set and the testing set are constructed based on the LIBS spectra acquired by Mars Surface Composition Detector (MarSCoDe) for China’s Tianwen-1 Mars exploration mission. All the spectra were collected from dedicated rock pellet samples, which were placed in a simulated Martian atmospheric environment. The classification performance of the CNN has been compared with that of three alternative machine learning algorithms, i.e., logistic regression (LR), support vector machine (SVM), and linear discriminant analysis (LDA). Among the four methods, it is on the CNN model that the highest classification correct rate has been obtained, as assessed by precision score, recall score, and the harmonic mean of precision and recall. Furthermore, the classification accuracy is inspected more quantitatively via Brier score, and the CNN is still the best performing model. The results demonstrate that the CNN-based chemometrics are an efficient tool for rock identification with LIBS spectra collected in a simulated Martian environment. Despite the relatively small sample set, this study implies that CNN-supported LIBS classification is a promising analytical technique for Tianwen-1 Mars mission and more planetary explorations in the future.
The Mars Surface Composition Detector (MarSCoDe) carried by the Zhurong rover of China’s Tianwen-1 mission uses Laser-Induced Breakdown Spectroscopy (LIBS) to detect and analyze the material composition on Martian surfaces. As one extraterrestrial remote LIBS system, it is necessary to adopt effective and reliable preprocessing methods to correct the spectral drift caused by the changes in environmental conditions, to ensure the analysis accuracy of LIBS scientific data. This paper focuses on the initial spectral drift correction and estimates the accuracy of on-board wavelength calibration on the LIBS calibration target measured by the MarSCoDe LIBS. There may be two cases during the instrument launch and landing, as well as the long-term operation: (a) the initial wavelength calibration relationship can still apply to the on-board LIBS measurement; and (b) the initial wavelength calibration relationship has been changed, and a new on-board calibration is needed to establish the current relationship. An approach of matching based on global iterative registration (MGR) is presented in respect to case (a). It is also compared with the approach of particle swarm optimization (PSO) for case (b). Furthermore, their accuracy is estimated with the comparison to the National Institute of Standards and Technology (NIST) database. The experimental results show that the proposed approach can effectively correct the drift of the on-board LIBS spectrum. The the root-mean-square error (RMSE) of the internal accord accuracy for three channels is 0.292, 0.223 and 0.247 pixels, respectively, compared with the corrected Ti-alloy spectrum and the NIST database, and the RMSE of the external accord accuracy is 0.232, 0.316 and 0.229 pixels, respectively, for other samples. The overall correction accuracy of the three channels is better than one-third of the sampling interval.
The Zhurong rover of China’s Tianwen-1 mission started its inspection tour on Mars in May 2021. As a major scientific payload onboard the Zhurong rover, the Mars Surface Composition Detector (MarSCoDe) instrument adopts laser-induced breakdown spectroscopy (LIBS) to detect and analyze the chemical composition of Martian materials. This paper introduces an experimental platform capable of establishing a simulated Martian atmospheric environment, in which a duplicate model of the MarSCoDe flight model is placed. In the simulated environment, the limit vacuum degree can reach 10−5 Pa level, the temperature can change from −190 °C to +180 °C, and different gases can be filled and mixed according to desired proportion. Moreover, the sample stage can move along a track inside the vacuum chamber, enabling the detection distance to vary from 1.5 m to 7 m. Preliminary experimental results indicate that this platform is able to simulate the scenario of MarSCoDe in situ LIBS detection on Mars well.
Quantum machine learning (QML) is an emerging discipline that combines quantum computing and machine learning and is able to exhibit exponential superiority over classical machine learning regarding computing speed on specific problems. This article provides a comprehensive review of the QML research in China. The QML development in China is presented in terms of research ideas and tasks, and the algorithms and application fields are sorted out. We have also highlighted some typical creative studies and illuminated their innovation points. Furthermore, the current challenges and future prospects are discussed. This review may provide inspiration for both China’s and global QML-domain progress.
Laser-induced breakdown spectroscopy (LIBS) has played a critical role in Mars exploration missions, substantially contributing to the geochemical analysis of Martian surface substances. However, the complex nonlinearity of LIBS processes can considerably limit the quantification accuracy of conventional LIBS chemometric methods. Hence chemometrics based on artificial neural network (ANN) algorithms have become increasingly popular in LIBS analysis due to their extraordinary ability in nonlinear feature modeling. The hidden layer activation functions are key to ANN model performance, yet common activation functions usually suffer from problems such as gradient vanishing (e.g., Sigmoid and Tanh) and dying neurons (e.g., ReLU). In this study, we propose a novel LIBS quantification method, named the Bayesian optimization-based tunable Softplus backpropagation neural network (BOTS-BPNN). Based on a dataset comprising 1800 LIBS spectra collected by a laboratory duplicate of the MarSCoDe instrument onboard the Zhurong Mars rover, we have revealed that a BPNN model adopting a tunable Softplus activation function can achieve higher prediction accuracy than BPNN models adopting other common activation functions if the tunable Softplus parameter β is properly selected. Moreover, the way to find the proper β value has also been investigated. We demonstrate that the Bayesian optimization method surpasses the traditional grid search method regarding both performance and efficiency. The BOTS-BPNN model also shows superior performance over other common machine learning models like random forest (RF). This work indicates the potential of BOTS-BPNN as an effective chemometric method for analyzing Mars in situ LIBS data and sheds light on the use of chemometrics for data analysis in future planetary explorations.