基于分数迟延估计的外辐射源雷达杂波相消算法

在利用外辐射源构成的收发分置连续波雷达系统中,直达波及多径杂波严重地影响了雷达对运动目标的探测性能.针对这一问题,通过对直达波及多径反射杂波的到达时间进行分数插值估计,提出了一种连续波雷达杂波相消算法.该算法利用参考接收通道的输出信号作多时延的分数内插估计来进行自适应杂波对消.基于实测数据的处理结果表明所提算法能获得平均45dB的检测增益,有效地改善了系统对运动目标的检测性能.

带纯迟延估计的过热汽温约束预测控制

针对发电机组过热汽温被控对象纯迟延、时变、不确定的特性,将带纯迟延估计的递推最小二乘算法引入输入受限的改进广义预测控制,提出了一种带纯迟延估计的过热汽温约束预测控制策略。系统内回路采用PID控制,内回路和主调节区对象构成广义被控对象,对广义被控对象进行拟合简化为二阶加纯迟延对象,以此作为本方法的被控对象。仿真结果表明,该方法的控制品质明显优于PID,具有较强的鲁棒性。尤其当对象纯迟延时间发生变化时,该方法比常规的改进广义预测控制取得更好的控制效果。

基于迟延估计与Kalman状态跟踪的热工过程动态数据驱动建模

针对常规热工对象历史数据建模过程中稳态数据获取困难的问题,从数据选取、数据处理和方法设计3方面出发,提出一种基于迟延估计与Kalman状态跟踪的热工过程动态数据驱动建模方法。该方法选取动态历史数据作为建模数据,将数据末端输入值作为输入稳态分量,将纯迟延时间及输出的稳态分量作为寻优变量的2个维度,并依据各量的值对数据进行处理;应用Kalman滤波算法获取系统初始状态后,结合智能寻优算法对系统进行建模仿真,对某火电机组高温过热器惰性区进行建模。结果表明:该建模方法能够直接应用系统的动态历史数据进行模型辨识,且所建模型具有较好的模型精度,为动态历史数据建模提供了参考。

基于IHHO-LSTM的SCR脱硝反应器出口NO_(x)浓度预测

针对燃煤电厂SCR脱硝反应器出口氮氧化物浓度测量存在较大迟延,以及吹扫阶段数据失真等问题,提出一种基于改进哈里斯鹰算法(Improved Harris hawk optimization,IHHO)和长短时记忆神经网络(Long short-term memory neural network,LSTM)的NOx浓度预测模型。首先利用最大信息系数对不同输入变量进行迟延时间估计,并通过最大相关最小冗余算法完成输入变量的选择并重构输入特征序列。然后,使用重构数据来完成LSTM模型的建立,并利用IHHO对网络相关参数进行优化。最后将所建模型的预测结果与未进行特征选择的IHHO-LSTM模型、进行特征选择后的HHO-LSTM模型和LSTM模型进行验证对比。结果表明,相较于其他预测模型,所提的IHHO-LSTM模型预测精度更高,具有较好的动态特性。

锅炉承压管泄漏双曲面定位的遗传算法优化

电站锅炉"四管"泄漏的精确定位对于火电机组的安全、经济运行具有极其重要的意义。该文建立了双曲面定位模型并给出炉膛内平面四元阵列的测点分布,泄漏定位实验表明通过极大似然(maximum likelihood,ML)广义互相关对能取得非常稳定和尖锐的时间估计峰值,频域信噪比函数能较好地抑制炉膛强背景噪声的影响。为了避免初值猜测与远场假设,采用遗传算法:十进制浮点编码,非均匀高斯变异算子,进行双曲面方程组定位解的全局性概率搜索。对比遗传算法和拟牛顿法的双曲面方程组优化,表明遗传算法的优化误差略小于拟牛顿法,具有稳健性和高精度,但其需要较长的收敛时间。时延估计误差对定位影响很大,时间迟延误差不大于0.1ms,平面四元阵列空间坐标定位误差保持在1m之内从而满足精确定位的要求。

基于DTV广播的外辐射源雷达对消算法研究

基于数字电视广播(DTV)的外辐射源探测系统,因为其抗隐身、反侦察、生存能力强和可靠性高等特点,成为现有探测技术的一种有效补充,备受各国重视。外辐射源探测系统中,影响运动目标探测性能的关键因素之一是直达波及多径杂波抑制。相对调频广播信号,数字电视广播信号带宽接近8MHz,地物杂波占据更多距离单元,常规的自适应对消算法的性能降低。针对这一问题,提出了分数阶自适应杂波相消的算法。基于实测数据的处理结果表明,该算法的对消性能明显改善,剩余杂波能量降低10dB以上。该算法已经应用于实际工程系统,取得了良好的效果。

Reference information time delay estimation and compensation in transfer alignment

To reduce the error in transfer alignment caused by reference information delay,a time delay estimation method is developed based on least-squares curve fitting of the angular rate integration.First,the gyro sensor measurements of the main strapdown inertial navigation system（M-SINS） and the slave strapdown inertial navigation system（S-SINS） are recorded for a few seconds and the integration of the data is calculated.Then,the possible maximum range of the delay value is defined and the points of the curve at different intervals are moved.The square of the differences between the corresponding points are calculated.Finally,the delay estimation can be acquired by the least-squares curve fitting of the M-SINS and the S-SINS.A delay compensation method by local data shifting is also presented.The simulation results demonstrate the effectiveness of delay estimation and indicate that the estimation accuracy is independent of the delay value.And the local data shifting compensation method can effectively reduce the errors of the transfer alignment caused by the reference information delay.

Characteristic Analysis of Underwater Acoustic Scattering Echoes in the Wavelet Transform Domain

Underwater acoustic scattering echoes have time–space structures and are aliasing in time and frequency domains. Different series of echoes properties are not identified when incident angle is unknown. This article investigates variations in target echoes of monostatic sonar to address this problem. The mother wavelet with similar structures has been proposed on the basis of preprocessing signal waveform using matched filter, and the theoretical expressions between delay factor and incident angle are derived in the wavelet domain. Analysis of simulation data and experimental results in free-field pool show that this method can effectively separate geometrical scattering components of target echoes. The time delay estimation obtained from geometrical echoes at a single angle is consistent with target geometrical features, which provides a basis for object recognition without angle information. The findings provide valuable insights for analyzing elastic scattering echoes in actual ocean environment.

Use of Delayed Channel Estimation Results in Adaptive Modulation

In this paper,potential use of perfect but delayed channel estimates for variable-power discrete-rate adaptive modulation is explored.Research is concentrated on block by block adaptation.At first,a new quantity-TAUD(Tolerable Average Use Delay)is defined,it quantifies the performance of an adaptation scheme in tolerating the delay of channel estimates.Then,the research on TAUD shows that the delay tolerating performance declines with the increase in average power,the scheme working with more modulation modes can tolerate a longer delay,and such improvement will be more significant with the increase of average power.Finally,it shows that,as the delay tolerating performance determines the maximum block length,it has a great effect on the maximum spectral efficiency.The criterion for determining the block length appropriate for the target BER is described and a simple method of calculating the maximum block length is presented.

BLIND CHANNEL ESTIMATION IN DELAY DIVERSITY FOR FREQUENCY SELECTIVE CHANNELS

Delay diversity is an effective transmit diversity technique to combat adverse effects of fading. Thus far, previous work in delay diversity assumed that perfect estimates of current channel fading conditions are available at the receiver and training symbols are required to estimate the channel from the transmitter to the receiver. However, increasing the number of the antennas increases the required training interval and reduces the available time with in whichdata may be transmitted. Learning the channel coefficients becomes increasingly difficult for the frequency selective channels. In this paper, with the subspace method and the delay character of delay diversity, a channel estimation method is proposed, which does not use training symbols. It addresses the transmit diversity for a frequency selective channel from a single carrier perspective in the form of a simple equivalent flat fading model. Monte Carlo simulations give the performance of channel estimation and the performance comparison of our channel-estimation-based detector with decision feedback equalization, which uses the perfect channel information.

Analysis of Traffic Parameter Estimation and Its Impacts on Wireless Channel

Wide band or broadband access was paid much attention with the development of radio transmission technique. The wireless access control procedure play an important role in this type of system and efficiency of control algorithm has a great impact on throughput of channel resource. Based on wide band network control model and the characteristics of radio channel, this paper proposed a channel traffic estimation method and then performed a dynamic parameter control procedure and give detail analysis on estimation error and its impact on channel throughput and delay performance. Computation and simulation of system performance show a positive solntion on system design.

Polynomial Root Finding on Frequency Estimation with Sub-Nyquist Temporal Sampling

This paper addresses an algebraic approach for wideband frequency estimation with sub-Nyquist temporal sampling. Firstly, an algorithm based on double polynomial root finding procedure to estimate aliasing frequencies and joint aliasing frequencies-time delay phases in multi-signal situation is presentcd. Since the sum of time delay phases determined from the least squares estimation shows the characteristics of the corre- sponding parameters pairs, then the pairmatching method is conducted by combining it with estimated parameters mentioned above. Although the proposed method is computationally simpler than the conventional schemes, simulation results show that it can approach optimum estimation performance.

Crustal structure of northeastern margin of the Tibetan Plateau by receiver function inversion

Using seismic data of about one year recorded by 18 broadband stations of ASCENT project, we obtained 2547 receiver func- tions in the northeastern Tibetan Plateau. The Moho depths under 14 stations were calculated by applying the H-x domain search algorithm. The Moho depths under the stations with lower signal-noise ratio （SNR） were estimated by the time delay of the PS conversion. Results show that the Moho depth varies in a range of -40--60 kin. The Moho near the Haiyuan fault is vague, and its depth is larger than those on its two sides. In the Qinling-Qilian Block, the Moho becomes shallower gradually from west to east. To the east of 105~E, the average depth of the Moho is 45 km, whereas the west is 50 km or even deeper. Combining our results with surface wave research, we suggest a boundary between the Qinling and the Qilian Mountains at around 105~E. S wave velocities beneath 15 stations have been obtained through a linear inversion by using Crust2.0 as an ini- tial model, and the crustal thickness that was derived by H-x domain search algorithm was also taken into account. The results are very similar to the results of previous active source studies. The resulting figure indicates that low velocity layers devel- oped in the middle and lower crust beneath the transition zone of the Tibet Block and western Qinling, which may be related to regional faults and deep earth dynamics. The velocity of the middle and lower crust increases from the Songpan Block to the northeastern margin of Tibetan Plateau. Based on the velocity of the crust, the distribution of the low velocity zone and the composition of the curst （Poisson＇s ratio）, we infer that the crust thickening results from the crust shortening along the direc- tion of compression.

Accuracy Improvement of Zenith Tropospheric Delay Estimation Based on GPS Precise Point Positioning Algorithm

In the precise point positioning(PPP),some impossible accurately simulated systematic errors still remained in the GPS observations and will inevitably degrade the precision of zenith tropospheric delay(ZTD) estimation.The stochastic models used in the GPS PPP mode are compared.In this paper,the research results show that the precision of PPP-derived ZTD can be obviously improved through selecting a suitable stochastic model for GPS measurements.Low-elevation observations can cover more troposphere information that can improve the estimation of ZTD.A new stochastic model based on satellite low elevation cosine square is presented.The results show that the stochastic model using satellite elevation-based cosine square function is better than previous stochastic models.

RISK-SENSITIVE FIXED-POINT SMOOTHING ESTIMATION FOR LINEAR DISCRETE-TIME SYSTEMS WITH MULTIPLE OUTPUT DELAYS

This paper investigates the risk-sensitive fixed-point smoothing estimation for hnear omcrete-time systems with multiple time-delay measurements. The problem considered can be converted into an optimization one in indefinite space. Then the risk-sensitive fixed-point smoother is obtained by solving the optimization problem via innovation analysis theory in indefinite space. Necessary and sufficient conditions guaranteeing the existence of the risk-sensitive smoother are also given when the risk-sensitive parameter is negative. Compared with the conventional approach, a significant advantage of presented approach is that it provides less computational cost.

Precise absolute positioning for a single-frequency user

A new strategy to realize precise absolute positioning for a single-frequency user is presented. In the presented strategy, the receiver clock and ambiguities are removed using the satelliteand epoch-differenced (SDED) algorithm. As a further development of the SDED algorithm, a regional augmentation network is used to generate the SDED atmospheric delays at the user. The weakened mathematic property due to the epoch-differenced operation is improved by adding the generated atmospheric delays and applying the robust estimation. To test the new approach, the 24-h data at 5 Continuous Operation Reference Station (CORS) stations in Shanghai is processed. The results show a more than 96% success rate, defined as the case where three directions achieve the desired positioning accuracy of 10 cm, when the observation is longer than 20 min. The 20-min static results show that the new method can reach an accuracy of 3.42, 4.76 and 9.26 cm in the North, East and Up directions, respectively. An experiment was carried out to assess the kinematic positioning. The results show that the kinematic positioning accuracy is 4.11, 5.31 and 4.05 cm in the north-south, east-west and height directions,respectively.