Table of Content

  

20 June 2017
Volume 47 Issue 3

Finite-time stochastic generalized outer synchronization of time-delayed complex dynamical networks

LI Wang, MA Zhicai, SHI Hongjun

JOURNAL OF SHANDONG UNIVERSITY (ENGINEERING SCIENCE). 2017, 47(3):  1-8.  doi:10.6040/j.issn.1672-3961.0.2016.320

Abstract ( 1332 )   PDF (1559KB) ( 585 )   Save

References | Related Articles | Metrics

Based on the finite-time control technology and the open-loop control method, the generalized outer synchronization between two complex dynamical networks with time delay and noise perturbation was investigated. A new finite-time controller was designed and the sufficient condition for the finite-time stochastic generalized outer synchronization was obtained based on the finite-time stability theory of stochastic differential equations. The results showed that the synchronization scheme was robust to the noise perturbation. The theoretical results showed that the synchronization time depended on the control strength. Under the same conditions, the synchronization time decreased with the increasing of the control strength. In the numerical examples, the R(¨overo)ssler-like system and Hindmarsh-Rose system were chosen as the node dynamics of the drive and response networks, respectively. The time evolution trajectories of synchronization error and total synchronization error were given. The effectiveness and feasibility of the theoretical result was confirmed by the numerical results.

Classification and analysis of epileptic EEG based on complex networks

HAO Chongqing, WANG Zhihong

JOURNAL OF SHANDONG UNIVERSITY (ENGINEERING SCIENCE). 2017, 47(3):  8-15.  doi:10.6040/j.issn.1672-3961.0.2016.279

Abstract ( 1473 )   PDF (5850KB) ( 734 )   Save

References | Related Articles | Metrics

To extract epileptic EEG features in the ictal and interictal period, a method of depicting epileptic EEG was proposed by transforming epileptic EEG time series to epileptic networks. Chaotic multi-dimensional time series coming from the Lorenz system and Rössler system were used to construct a complex network,in which all the variables could be measured. It was found that there was morphological similarity between topology of the complex networks and the attractor of chaotic system. This indicated that complex networks constructed from time series could depict the characteristics of the original signals. For only one measureable variable, multi-dimensional time series were obtained by reconstruction of the phase space. Therefore, the epileptic EEG network was constructed and analyzed in the ictal and interictal period. The results showed that epileptic EEG network topologies in the ictal period were significantly different from that in the interictal period. Meanwhile, the average path length of the network increased significantly and recurrence rates decreased significantly in the ictal period comparing to in the interictal period. These network features could be used to depict the characteristics of EEG time series and could provide the basis for epilepsy automatic identification and prediction.

Image patch prior based denoising algorithm by using low rank approximation and Wiener filtering

ZHANG Yang, CHEN Fei, XU Haiping

JOURNAL OF SHANDONG UNIVERSITY (ENGINEERING SCIENCE). 2017, 47(3):  16-20.  doi:10.6040/j.issn.1672-3961.0.2017.009

Abstract ( 1436 )   PDF (1499KB) ( 745 )   Save

References | Related Articles | Metrics

A Gaussian mixture model(GMM)was used to study the texture structure of natural image patches, and a low-rank approximation and Wiener filtering algorithm based on image patch prior were proposed. The proposed method divided the image into a number of overlapped patches and clustered them for collaborative filtering by using the prior structures of external image patch and internal image self-similarity. By grouping nonlocal similar patches, low-rank approximation was used as collaborative filtering to recover the texture structures. When the number of similar patches was small, Wiener filtering with patch prior was adopted to preserve texture features. The experimental results indicated that the proposed method was more suitable for the images with fewer similar patches like boundary and corner etc., and showed very competitive performance with state-of-the-art denoising method in terms of Peak Signal to Noise Ratio(PSNR)and visual quality.

Research on the classification of resting state EEG signal between depression patients and normal controls by EMD and SVM methods

LIU Yan, LI Youjun, CHEN Meng

JOURNAL OF SHANDONG UNIVERSITY (ENGINEERING SCIENCE). 2017, 47(3):  21-26.  doi:10.6040/j.issn.1672-3961.1.2016.148

Abstract ( 1684 )   PDF (1886KB) ( 696 )   Save

References | Related Articles | Metrics

Automatic detection of depression state was significant for mental disease diagnostics and rehabilitation, which could decrease the duration of work required when inspecting the electroencephalography(EEG)signals. A novel method for feature extraction and pattern recognition from subjects resting state EEG signal, based upon empirical mode decomposition(EMD)and support vector machine(SVM)was proposed to make a distinction between depression patients and normal controls. The EEG signals were collected from 20 depression patients and 25 normal persons, and the EEG was filtered and extracted as features. The SVM was used as classifier for recognition which showed whether the person was a depression patient. The experimental results showed that the algorithm could achieve the specificity of 93.3%. And the classification accuracy from the features extracted by EMD was higher than the classification accuracy from features extracted by wavelet clearly.

Visual saliency detection based on visual center shift

HU Jinge, TANG Yan

JOURNAL OF SHANDONG UNIVERSITY (ENGINEERING SCIENCE). 2017, 47(3):  27-33.  doi:10.6040/j.issn.1672-3961.0.2016.340

Abstract ( 1349 )   PDF (3343KB) ( 451 )   Save

References | Related Articles | Metrics

Many existing detection methods could not extract saliency regions clearly. A novel saliency detection method based on visual center offsetting was proposed. On the basis of images' pre-segmentation, combining with color contrast features, color distribution features and location features, saliency region of an image was extracted. The center offsetting was used to simulate the vision transfer process of human, after multi-scale analysis, by fusing saliency maps at different scales. The final saliency map was computed. The results showed that the performance of the proposed method was better on visual effect and the precision recall rate than existing methods, the area under ROC curve was 0.952.

A feature selection method based on LS-SVM and fuzzy supplementary criterion

LI Sushu, WANG Shitong, LI Tao

JOURNAL OF SHANDONG UNIVERSITY (ENGINEERING SCIENCE). 2017, 47(3):  34-42.  doi:10.6040/j.issn.1672-3961.0.2016.308

Abstract ( 1499 )   PDF (1576KB) ( 476 )   Save

References | Related Articles | Metrics

Traditional feature selection algorithm used a single scalar metric such that it might become difficult to achieve a trade-off between generalization performance and dimension reduction at the same time. A new feature selection algorithm called LS-SVM-FSC was proposed to circumvent this shortcoming. The kernel-based least squares support vector machines was used to train a set of binary classifiers on each single feature and a kind of new fuzzy membership function was used to obtain fuzzy membership value of each pattern belonging to its class. Based on a new fuzzy supplementary criterion, the features with minimal redundancy and maximal relevance was selected. Experiments indicated that the proposed algorithm had high classification accuracy and strong dimension reduction capability on nine datasets. In particular, it still kept fast learning speed for high-dimensional datasets, in contrast to other ten feature selection methods and seven degree determination methods.

The video synopsis based on the enhanced ViBe algorithm

HUI Kaifa, CHENG Keyang, ZHAN Yongzhao

JOURNAL OF SHANDONG UNIVERSITY (ENGINEERING SCIENCE). 2017, 47(3):  43-48.  doi:10.6040/j.issn.1672-3961.0.2016.306

Abstract ( 1506 )   PDF (2017KB) ( 410 )   Save

References | Related Articles | Metrics

Focusing on the time redundancy of surveillance video, an enhanced ViBe was proposed to solve the problems of noise and the ghost in ViBe algorithm. The improved algorithm was applied in the procession of video background modelling. It could be determined whether there was a foreground object in a certain frame by extracting outside contour of the obtained binary image, and the frames contains foreground objects would be pushed into the video stream for the purpose of video synopsis. After the experimental verification, it could be concluded that the method could effectively reduce the redundant information in the video and the volume of the video. Meanwhile some important information in the video could be retained, and the algorithm satisfied the requirement of real-time.

Brain MR image segmentation based on student's t mixture model with Markov random field

LI Lu, FAN Wentao, DU Jixiang

JOURNAL OF SHANDONG UNIVERSITY (ENGINEERING SCIENCE). 2017, 47(3):  49-55.  doi:10.6040/j.issn.1672-3961.0.2016.310

Abstract ( 1344 )   PDF (1581KB) ( 546 )   Save

References | Related Articles | Metrics

Extra computation was always needed when using Expectation Maximization(EM)algorithm for solving mixture models. To overcome this drawback, a novel Student's t-mixture model based on Markov random field was proposed. EM algorithm was used directly in the proposed model, which was convenient and efficient. According to the experimental results, the proposed method could overcome the impact of noise on the segmentation results efficiently, and got better segmentation results.

An image saliency object detection algorithm based on adaptive manifold similarity

REN Yongfeng, DONG Xueyu

JOURNAL OF SHANDONG UNIVERSITY (ENGINEERING SCIENCE). 2017, 47(3):  56-62.  doi:10.6040/j.issn.1672-3961.0.2016.305

Abstract ( 1357 )   PDF (2813KB) ( 737 )   Save

References | Related Articles | Metrics

In order to improve the adaptability and precision in extracting salient regions in images, an image salient region detection algorithm was proposed based on adaptive manifold similarity. An input image was segmented into super-pixels which were represented as the nodes in a graph. The node with high frequency was generated by the characteristics of the salient regions. Convex hull computation was used to generate the saliency seeds of the salient object area according to high-frequency nodes. The proposed algorithm was used to complete information reconstruction of the current image by adaptively assessing the salient weights on the edges between the nodes. In addition, based on local characteristics information reconstruction, the proposed algorithm utilized similarity extraction function to self-adaptively obtain the similarity characteristics and manifold structures in order to spread salient characteristics information. The experimental results showed that the quadratic programming solution exploited to compute the weights between the nodes could effectively avoid threshold selection and enhance robustness accordingly, and the proposed method performed better than the other state-of-the-art methods.

Surgical navigation system based on anartificialmarker

MA Shuaiyifan, ZHAO Zijian

JOURNAL OF SHANDONG UNIVERSITY (ENGINEERING SCIENCE). 2017, 47(3):  63-68.  doi:10.6040/j.issn.1672-3961.0.2017.007

Abstract ( 1531 )   PDF (3588KB) ( 734 )   Save

References | Related Articles | Metrics

In order to solve the shortcomings of the current surgical navigation systems including weak anti-interference ability, complex system and non-multi-target tracking, an implement of surgical navigation system based on the detecting of artificial markers was proposed. Based on the Bumblebee2 binocular vision system, a detection algorithm based on edge detection was proposed to realize the precise and fast detection of ARTag artificial marks by line detection, marker construction and marker decode. Then with the Binocular stereo vision theory, accurately and reliably calculatedspatial location of the surgical instruments. Also, the system was able to tracking multiple targets in real-time because these artificial markers were set number sequences inside which enable the system to detect, decode and match multiple artificial marker targets. Experimental results showed that the proposed algorithm was of high accuracy and strong anti-interference ability, and the navigation system solution was fast, stable and economical, advantages multiple targets tracking at the same time, had broad prospects in actual use.

Survey on medical robot in computer-aided surgery

ZHAO Zijian, WANG Fang, CHANG Faliang

JOURNAL OF SHANDONG UNIVERSITY (ENGINEERING SCIENCE). 2017, 47(3):  69-78.  doi:10.6040/j.issn.1672-3961.0.2016.482

Abstract ( 1465 )   PDF (2000KB) ( 1131 )   Save

References | Related Articles | Metrics

The application of medical robot technique in computer-aided surgery was a hot research topic at present. The concept of medical robot was introduced, and the necessity of its application in computer-aided surgery was also emphasized. The state of arts of medical robot technique was discussed systematically in computer-aided surgery, covering the four aspects: orthopedic surgery, neurological surgery, laparoscopic surgery and intervention surgery. Besides, some corresponding medical robot systems were also introduced. Through the analysis on the medical robot technique, the laparoscopic robot was considered as the most promising research direction. There were four urgent problems concerning the laparoscopic robot. Based on the four problems, further research questions and possible directions in the future were proposed.

Sliding model chaos synchronization control of a class of fractional-order multi-scroll systems

MAO Beixing, WANG Dongxiao

JOURNAL OF SHANDONG UNIVERSITY (ENGINEERING SCIENCE). 2017, 47(3):  79-83.  doi:10.6040/j.issn.1672-3961.0.2016.058

Abstract ( 1304 )   PDF (1449KB) ( 460 )   Save

References | Related Articles | Metrics

The problem of sliding model chaos synchronization of a class of fractional-order multi-scroll systems with certain and uncertain parameter was studied based on fractional order calculus theory and sliding mode control approach. The switching function and controller was designed, two sufficient conditions were arrived for the fractional order systems sliding model synchronization. The research conclusion illustrated that fractional-order multi-scroll systems was sliding mode chaos synchronization under proper controllers and self-adaptive law.

Terminal sliding model control chaos synchronization of fractional-order emotion mode systems

LI Qingbin, WANG Xiaodong

JOURNAL OF SHANDONG UNIVERSITY (ENGINEERING SCIENCE). 2017, 47(3):  84-88.  doi:10.6040/j.issn.1672-3961.0.2016.342

Abstract ( 1373 )   PDF (1083KB) ( 347 )   Save

References | Related Articles | Metrics

The problem of terminal sliding model synchronization of fractional-order emotion mode systems was studied based on drive-response approach. Based on Lyapunov stability theory and fractional-order calculus theory, nonsingular sliding mode surface was designed. A designing project for the master-slave systems getting fast speed synchronization in finite-time was given by designing terminal sliding model controllers. Both the theoretical analysis and simulation results illustrated the effectiveness of this control method.

Feature extraction method of rolling bearing inner ring in wind turbine based on improved EMD and feature box

YU Qingmin, LI Xiaolei, ZHAI Yong

JOURNAL OF SHANDONG UNIVERSITY (ENGINEERING SCIENCE). 2017, 47(3):  89-95.  doi:10.6040/j.issn.1672-3961.0.2016.270

Abstract ( 1520 )   PDF (3335KB) ( 552 )   Save

References | Related Articles | Metrics

According to the characteristics of vibration signal of rolling bearing inner ring in direct-driven wind turbine, a new method of fault diagnosis by improved empirical mode decomposition(EMD)and feature box was put forward. The original signal was decomposed by improved EMD to get a finite number of stationary intrinsic mode functions(IMFs). The characteristics of amplitude domain parameters such as mean and variance were extracted, which were turned into feature matrix chose by effectiveness. To perform data smoothing processing, The feature matrix was divided into boxes and replaced by means of data in each box. Examples showed that the feature matrix, which was divided into boxes finally, could effectively extract the fault feature of rolling bearing, and reduce the over fitting of the machine learning model.

A position sensorless control strategy for BLDCM based on a fractional order sliding mode observer

LIANG Qiushi, ZHAO Zhicheng

JOURNAL OF SHANDONG UNIVERSITY (ENGINEERING SCIENCE). 2017, 47(3):  96-101.  doi:10.6040/j.issn.1672-3961.0.2016.298

Abstract ( 1436 )   PDF (4397KB) ( 587 )   Save

References | Related Articles | Metrics

In order to solve the problem of chattering and the phase delay in traditional sliding mode observer, the fractional order mode was introduced to the sliding mode observer design procedure. Combined with brushless DC motor(BLDCM)line voltage equation, the proposed sliding mode observer was designed by utilizing the slower energy transferring property of fractional order, which avoided observed phase delay in traditional integral order observer as a result of low pass filter existence. At the same time, the smooth assessment curve of line back electromotive force could be obtained, and the real-time and accurate rotor position signal and speed assessment value in this scenario could further be obtained. The simulation result indicated that the proposed sliding mode observer was able to observe the line back electromotive force of BLDCM accurately and achieve its position sensorless control.

Risk control of extremely close twin shield tunnels nearby passing intensive buildings based on informative construction

LIU Jifeng, YANG Huanhuan

JOURNAL OF SHANDONG UNIVERSITY (ENGINEERING SCIENCE). 2017, 47(3):  102-111.  doi:10.6040/j.issn.1672-3961.0.2016.276

Abstract ( 1280 )   PDF (5066KB) ( 445 )   Save

References | Related Articles | Metrics

To reduce the building subsidence induced by nearby passing twin shield tunnels, the FLAC^3D was adopted to numerical simulate and analyze the different shield tunnel excavation methods induced the closest building ground settlement. The shield tunneling parameters were optimized for adopted construction method that changed the right line direction with internal cross bracing in left line, and with twin lines reinforcement by tube grouting in surrounding ground. And the related site surveying parameters were presented and analyzed. The results showed that: the previous construction method induced the building differential settlement was 26.5 mm, which was severely beyond the 10 mm limit; piling between buildings and tunnels or internal cross bracing in left line and twin lines tube grouting in surrounding ground were all applicable reinforcement methods to meet 10 mm limit of building differential settlement; it reduced effectively construction induced building settlement by optimizing right line tunneling parameters; the main parameters of later excavated tunnel influence on previous constructed tunnel were tunneling load and grouting pressure, the use of real-time monitoring and information construction was necessary. The extremely close twin shield tunnel successfully nearby passed the buildings, and the results could be consulted for similar projects.

Seismic performance of RC pier with reinforcement corrosion

ZHAO Jianfeng, LI Hongyi, LIU Suwen

JOURNAL OF SHANDONG UNIVERSITY (ENGINEERING SCIENCE). 2017, 47(3):  112-118.  doi:10.6040/j.issn.1672-3961.0.2016.356

Abstract ( 1668 )   PDF (3096KB) ( 527 )   Save

References | Related Articles | Metrics

In order to study the seismic performance of the reinforced concrete bridge piers in service, which was affected by the reinforcement corrosion due to the chloride ion corrosion, a cylindrical pier was taken as an example. By means of probability method, the initial corrosion time and the corrosion rate were determined, the deterioration models of the diameter and the yield strength of steel were established. The nonlinear dynamic analysis model was made by ABAQUS software and three analysis cases as consisting of stirrup corrosion, longitudinal reinforcement corrosion and the combination corrosion of deterioration, were investigated to study the seismic performance of reinforced concrete piers caused by material deterioration. The results showed that compared with only considering the corrosion situation of longitudinal reinforcement in the past, the deterioration of seismic performance was more serious when the stirrups and longitudinal reinforcement corrosion were both considered, the ductility performance of the bridge pier was affected seriously by the stirrup corrosion and the bending performance of the bridge pier was affected seriously by the longitudinal reinforcement corrosion.

The influence of grouting pressure and diffusion features of different grouting materials

PENG Rongrong, YU Wensheng

JOURNAL OF SHANDONG UNIVERSITY (ENGINEERING SCIENCE). 2017, 47(3):  119-124.  doi:10.6040/j.issn.1672-3961.0.2016.383

Abstract ( 1630 )   PDF (1980KB) ( 1049 )   Save

References | Related Articles | Metrics

A new grouting simulation device was developed to study the influence of property of grout slurry on grouting pressure and slurry diffusion. This device substituted overall filling and compaction of soil of layer-compacting to avoid the occurring of interlinear interface. Thus the influence of interlinear interface on slurry diffusion could be decreased. Grouting pressure and slurry diffusion of sodium carboxymethyl cellulose solution, simulating Newtonian slurry, cement slurry and cement-silicate slurry were studied by simulation experiment, and then validated by theoretical analysis. The results showed that property of grout slurry had remarkable influences on change laws of grouting pressure and splitting feature. Slurry diffusion pattern could be abstained by analyzing the change laws of grouting pressure in combination with splitting feature.

Silt mudstone stress relaxation characteristics under tri-axial stress condition

CHEN Fangming, HU Quanguang, NING Guangzhong

JOURNAL OF SHANDONG UNIVERSITY (ENGINEERING SCIENCE). 2017, 47(3):  125-129.  doi:10.6040/j.issn.1672-3961.0.2017.099

Abstract ( 1550 )   PDF (2071KB) ( 384 )   Save

References | Related Articles | Metrics

To investigate the stress relaxation characteristics of soft rock, step loading relaxation tests under tri-axial stress condition were performed with silt mudstone samples. Based on the analysis of silt mudstone relaxation characteristics with different confining stresses, the stress relaxation properties could be well reflected by the Burgers model. The evolution of stress relaxation rate with the change of original stresses was studied. The results showed that, when the original stress was lower, the stress relaxation rate decreased exponentially with the increase of the initial stress; while higher original stress was applied, the stress relaxation of some samples showed a slight increase trend.

Sharing mechanism of machine tool resource based on ontology

LI Qingdong, LUO Weichao, YE Yingxin, ZHANG Chengrui, HU Tianliang

JOURNAL OF SHANDONG UNIVERSITY (ENGINEERING SCIENCE). 2017, 47(3):  130-138.  doi:10.6040/j.issn.1672-3961.0.2016.405

Abstract ( 1343 )   PDF (4241KB) ( 439 )   Save

References | Related Articles | Metrics

In order to solve the key problem which existed in resource expression, storage and processing task matching aspects of machine tool resource sharing, a sharing mechanism of machine tool resource based on ontology was given. Ontology was used to express machine tool in semantic, which increased the readability by both human and machine. Distributed database HBase was applied to store machine tool ontology, tables were well designed for the purpose of efficient access. A machine tool resource query mechanism based on machining task was proposed to realize effective matching between machining tasks and sharing resources. After storing mass machine tool instances in the data base, a case study was researched through submitting a turning task to the resourcing sharing platform and finally a list of suitable lathes were retrieved. The result shows that the machine tool resource storage and query mechanism based on ontology can meet the needs of resource sharing,and the feasibility and availability of sharing mechanism were also verified.

Influence of different acid-etching time on the surface morphology and corrosion resistance of TC4 titanium alloys after sandblasting

REN Bing, WAN Yi, WANG Guisen, WANG Teng

JOURNAL OF SHANDONG UNIVERSITY (ENGINEERING SCIENCE). 2017, 47(3):  139-142.  doi:10.6040/j.issn.1672-3961.0.2017.003

Abstract ( 2087 )   PDF (2464KB) ( 1231 )   Save

References | Related Articles | Metrics

TC4 Titanium plates were etched by HCl/H₂SO₄ mixture after being blasted by Al₂O₃ particles of size 220 to investigate the effect of different etching time(0.5,1, 2 and 3 h)on the surface morphology and corrosion resistance, while the polishing treatment was set as the control group. The surface morphology of the titanium alloy samples was characterized by scanning electron microscopy(SEM)and 3D laser confocal microscopy followed by the surface roughness measurement. Then the corrosion resistance measurement was carried out by the electrochemical workstation to evaluate the corrosion behavior of all samples. The results showed that with the prolongation of etching time, the surface morphology of titanium alloy tended to be homogenized from clutter, and the roughness first rose and then decreased. Group 2 h had the maximum roughness and the highest proportion of surface micropores, and the interface expansion area ratio was 222%. In terms of corrosion resistance, all of the sandblasting and acid-etching(SLA)groups performed better than the polishing group, indicating that the SLA treatment can be beneficial for generating oxide film which possesses better corrosion resistance. Besides, with the extending of etching time, the corrosion tendency was smaller, and the group 3 h performed the best of all groups.

Preparation and optimization of compound bonded material used in reducing the contact thermal resistance

DONG Zhen, YANG Yonglu, XIONG Guodong, LAI Yanhua, LYU Mingxin

JOURNAL OF SHANDONG UNIVERSITY (ENGINEERING SCIENCE). 2017, 47(3):  143-150.  doi:10.6040/j.issn.1672-3961.0.2017.030

Abstract ( 1385 )   PDF (2024KB) ( 406 )   Save

References | Related Articles | Metrics

To enhance the global thermal conductivity of adsorbed bed, heating resistingand corrosion-resistant composite bonded material with high thermal conductivity was studied. It could be daubed between adsorbed bed and adsorbent for decreasing thermal contact resistance.The article experimental researched the influence of thermal conducting filler, diluent and coupling reagent and other factors on composite bonded materials thermal conductivity at first. And then confecting project of composite bonded material with orthogonal experiment was optimized. The research results showed that the thermal properties of the compositebonded material improved with the percentage of thethermal conducting filler Al₂O₃ increasing, composites with 10 μm Al₂O₃ fillers had better thermal performance than the composites with 35 μm Al₂O₃fillers.For 10 μm Al₂O₃ fillers, which treated by 8% concentration coupling agent had the highest thermal conductivity.The thermal conductivity of the composites could be significantly improved by using the appropriate ratio of the thermal conductive fillers, diluents and coupling agents.

Forecasting method of boiler heat load based on Markov chain

LU Changhai

JOURNAL OF SHANDONG UNIVERSITY (ENGINEERING SCIENCE). 2017, 47(3):  151-158.  doi:10.6040/j.issn.1672-3961.0.2017.129

Abstract ( 1384 )   PDF (1399KB) ( 352 )   Save

References | Related Articles | Metrics

The multi-parameter forecasting control model was proposed based on Markov chain of the forecasting method of coal-fired boiler heat load. In accordance with outdoor temperature, the average feed-water temperature was predicted. Then the compound probability was forecast with reference to the average daily fuel consumption, air-blower frequency and induced draft fan frequency. Conversely, the range of high probability was taken full advantage of to optimize the above mentioned three parameters.The experiment results showed that when the room temperature met the heating standards, the average daily fuel consumption was reduced by about 8 percent and the average daily feed-water about 7 percent after the forecasting control model was carried out,the accuracy of coal burning and operation control promoted effectively.This method provided a new way of the fine control of boil heat load.