This paper presents a vision/radar/INS integrated guidance system for shipboard landing. The vision system on the aircraft tracks landmarks on the ship and outputs their coordinates in the image. The motion information of aircraft is obtained from airborne Inertial Navigation System (INS). Meanwhile, the datalink receives the relative position between the ship and aircraft from the shipboard radar and the ship motion information from shipboard sensors. The federated filter is utilized to fuse the vision, radar and inertial information, and the information from INS is chosen as the reference system in the federated filter. The estimated ship motion information is utilized to calculate the glide-down path. All the guidance information is regarded as the feedback information given to the flight control system to calculate control command. A designed on-vehicle experiment shows that the vision/radar/INS integrated landing guidance system achieves satisfied results.

Rotating modulation technique is a mature method which has been widely used in the rotational inertial navigation system (RINS). Inertial Measurement Unit (IMU) can be driven to rotate to modulate the inertial devices’ errors, in this way the position accuracy of RINS can be greatly improved. However, the outputs of attitude are easily affected for the existence of rotation mechanism. The rotation mechanism of tri-axis RINS consists of three pairs of motors and photoelectric encoders. The errors of the rotation mechanism mainly contain the installation errors of the gimbals, the installation errors of photoelectric encoder and the zero-position errors of the system; they have different effects on the outputs of the attitude. This paper researches the principle of attitude output accuracy loss in tri-axis RINS, the effects of different errors on attitude are analyzed and a self-calibration method based on recursive least squares (RLS) is proposed to calibrate these errors. Simulation and experiment results prove that the analyses is right and experiment results show that the proposed attitude compensation method can improve short-term attitude accuracy greatly, after the compensation, the attitude accuracy of the rotation direction is less than 12” , and the attitude accuracy of the other direction is less than 8” .

The rotational inertial navigation system (RINS) could greatly improve navigation accuracy by rotating the inertial measurement unit with gimbals. But it will also excite and amplify the corresponding error parameters, so it is more necessary to calibrate the typical error parameters accurately. Fortunately, the self-calibration strategy could be achieved conveniently in RINS. In this paper, an innovative self-calibration strategy for error parameters of dual-axis is proposed. The calibration strategy is designed that the inner and outer gimbals rotate at different angular speeds at the same time, and the whole calibration process lasts for 12 min. The relationship between the error parameters to be calibrated and the navigation error is derived, and the principle of error separation is analyzed, which proves the rationality of the strategy. The calibration strategy is verified by simulation and experiment, and the 21 error parameters of the system can be well estimated. The 3h static navigation experiment shows that the position error is less than 0.15nmile/h(CEP) after compensating the error parameters calibrated by the self-calibration strategy, and achieved the expected accuracy of the dual-axis RINS. Finally, it is proved that the calibration strategy has the obvious advantages of short calibration time and simple process, which is of great significance to improve the navigation accuracy of dual-axis RINS.

Ship airwake is an important factor affecting the accurate carrier landing of carrier-based aircraft. To suppress the effects from airwake, this paper presents a flight control method using neural network prediction. Different from most studies focusing on improving the adaptability of the flight control methods, the emphasis of the research is placed on modelling the effects of the ship airwake in the carrier landing environment pertinently and reflecting the effects in the control model. From a large amount of carrier landing flight data, the impact of airwake on the landing control can be extracted and summarized into models based on neural network algorithm. Then the airwake impact in the next period of time can be predicted. A flight control law is designed combining the predicting model of airwake impact with model predictive control. The simulation demonstrated that the presented method can reduce the trajectory deviation influenced by airwake and decrease the touchdown dispersions.

In the rotational inertial navigation system (RINS), the horizontal gyro drift can be modulated when the IMU rotates around the z axis. At the same time, the scale factor error of the z gyro can be offset by forward and reverse rotation. However, in the forward and backward rotation, the gyro scale factor error will still cause the heading angle output error. Moreover, the change of the scale factor error of the fiber optic gyros (FOG) is sensitive to temperature changes, which makes it difficult to calibrate gyro scale factor error accurately. Therefore, in this paper, a method to calibrate the scale factor errors of two RINSs dynamically is proposed. During the movement of vehicle, the heading angle difference between the RINS1 and RINS2 is calculated real time, the scale factor errors of RINS1 and RINS2 could be estimated at the same time with the least squares filtering algorithm. The simulation experiments show that when RINS1 and RINS2 adopt a reasonable rotation strategy, the gyro scale factor errors were less than 0.5ppm. In the vehicle experiment, the heading angle error caused by the gyro scale factor error is kept below 5". And the method significantly improves the problem that scale factor error affected by temperature changes during navigation process, and at the same time solves the problem of the calibration coefficient error in the dynamic situation.

The ship usually adopts multiple sets of rotational modulation inertial navigation system (RINS) to ensure the accuracy and reliability of long endurance navigation, but the rotational modulation will excite the gyro installation error and scale factor error, which will cause periodic oscillation of navigation attitude information, moreover, due to the influence of temperature, vibration and other environments, these errors will change in the results of pre-calibration, and it is impossible to accurately model them, resulting in the further aggravation of attitudes oscillation, which is not conducive to the transfer alignment between master and slave INS. However, the ship is always swinging or maneuvering, a single RINS can not complete self-calibration independently in dynamic condition. In order to solve the above problems, in this paper, an online calibration algorithm based on joint rotation modulation of multi-RINSs is proposed for redundant configuration of multi-RINSs without extra conditions, and the error parameters are tracked and estimated every 4 minutes, the dynamic simulation results show that the calibration strategy can effectively calibrate the gyro installation error, the nonorthogonal angle between the gimbals and the scale factor error of the z-gyro of each RINS, which proves the correctness of the calibration strategy. The results of 3h navigation experiment show that the amplitude of attitudes oscillation decreases about 80% after the compensation of joint calibration error parameters, the fluctuation amplitude of heading angle decreases by 70%, the effectiveness of the calibration strategy in practical application is further verified, the relationship of ship multi-RINSs changes from backup to assistance, which is of great significance to improve the heading and attitudes accuracy of RINSs and the transfer alignment performance between master and slave INS.