A Bayesian posterior distribution-based abnormal data processing GM(0, N) model is proposed for complex equipment cost estimation to enhance the accuracy and reliability of cost prediction for sophisticated systems. Taking rocket systems as a case study, potential abnormal data samples were first eliminated using the three-sigma criterion to improve dataset quality and model precision. Subsequently, Bayesian estimation methods were employed to calculate the mean and variance of critical development cost-per-unit parameters (e.g., takeoff mass), providing a scientific foundation for subsequent outlier detection. A fractional-order accumulation GM (0, N) model was subsequently constructed, where the unknown parameters were determined through the least squares method, enabling precise development cost estimation for target rocket configurations. Experimental results demonstrate that compared with multivariate linear regression models and conventional GM(0, N) models, the proposed method achieves superior estimation accuracy and enhanced robustness. Specifically, the prediction error of this methodology was reduced to 2.637 5%, whereas the errors of multivariate linear regression and conventional GM (0, N) models reached 20.716 9% and 14.212 8% respectively. Furthermore, this methodology not only applies to liquid rocket development cost estimation but can also be extended to cost prediction problems in other complex engineering systems, providing valuable references for scientific research and technological development in related fields.

This study can play an important role in addressing the issue of achieving stable control of launch vehicles when the first-order elastic frequency is near or overlaps with the rigid-body frequency, especially without prior frequency knowledge. Much of the research up to now, such as those based on notch-filter or signal-processing techniques, requires a gap of 3 to 5 times between the low-order elastic frequencies and the rigid-body cutoff frequency. Observer-based methods, on the other hand, require prior knowledge of the frequency data. In this paper, a method is proposed, which takes advantage of the fact that attitude measurements of the rigid-body, obtained by Inertial Measurement Units (IMUs) installed at different locations, are identical, whereas elastic-mode signals differ only in their modal slopes. By differentially processing these IMU measurements, we can extract only the elastic information, thus identifying the frequencies of the elastic modes, and design an observer to distinguish the first-order elastic signal using redundant measurements. Simulations demonstrate that this method can autonomously identify low-order elastic frequencies, even when they are identical to or less than the rigid-body frequency, thereby enhancing controller performance.

Space transportation systems have supported China in completing numerous major space engineering projects, including manned spaceflight and lunar exploration missions. Facing the intense demands for future launches and the necessity of scheduled space transportation, there is a growing need for launch vehicles with higher development efficiency, greater adaptability, and improved comprehensive performance. Control technologies can play a significant role in this development process, and continuously innovating at the theoretical method level is a very cost-effective solution compared to other technical approaches. The paper highlights three key challenges in space transportation system control technology: the integrated design of guidance and trajectory planning, strong adaptive control in the face of global uncertainty, and active flow pattern control in cryogenic fluid transmission. For the first two technologies, their conceptual connotations, necessity, research and application status in the Long March launch vehicles, and future prospects are discussed. For the third technology, a technical roadmap for autonomous control based on multidisciplinary crossover solutions is proposed. These challenges have long been neglected due to the lack of good solutions, limiting the further improvement of launch vehicle performance. However, with these technological breakthroughs, we anticipate a significant advancement in space transportation systems.

In recent years, Chinese Long March (LM) launchers have experienced several launch failures, most of which occurred in their propulsion systems, and this paper studies Autonomous Mission Reconstruction (AMRC) technology to alleviate losses due to these failures. The status of the techniques related to AMRC, including trajectory and mission planning, guidance methods, and fault tolerant technologies, are reviewed, and their features are compared, which reflect the challenges faced by AMRC technology. After a brief introduction about the failure modes of engines that can occur during flight, and the fundamentals of trajectory planning and joint optimization of the target orbit and flight path, an AMRC algorithm is proposed for geostationary transfer orbit launch missions. The algorithm evaluates the residual performance onboard, and plans new objectives and corresponding flight path by iterative guidance mode or segmented state triggered optimization methods in real-time. Three failure scenarios that have occurred during previous LM missions are simulated to check the robustness of the algorithm: imminent explosion risk of the boosters’ engines, thrust drop during the first stage of flight, and being unable to start the engine during the second stage. The payloads would fall from space according to the current design under these conditions, but they were saved with the AMRC algorithm in the simulations, which allowed the rocket to get into the target orbit as intended or the payloads were deployed in other orbits without crashing. Although spaceflight can be very unforgiving, the AMRC algorithm has the potential to avoid the total loss of a launch mission when faced with these kinds of typical failures.