Search results

Then principal component analysis was used for data reduction and finally quadratic analysis was used to grade the apples [4].
They can be viewed as a family of generalized linear classifiers which find the desired function based on given empirical data.
Thus it can maximize predictive accuracy while automatically avoiding over-fit to the training data.
The smaller the pairwise misranking error is, the better the prediction fits the training data.
Perhaps the data set is too small.

Tab.1 The data of tank test The landing AUV II The landing AUV I Velocity(m/s) Re(E+6) X(N) Velocity(m/s) Re(E+6) X(N) 0.5 2.03 5.422 0.5 1.65 8.775 1.0 4.06 22.843 1.0 3.30 35.296 1.5 6.09 46.879 1.5 4.95 74.089 2.0 8.13 80.269 2.0 6.60 134.485 2.5 10.16 124.505 2.5 8.25 207.206 The dynamic calculation of the landing AUV II Dynamic stability calculation.
However, the ADCP (Acoustic Doppler Current Profiler), which presents velocity data for the underwater guidance system, will not work properly when the pitch angle is over 30 degrees.
The data of bottom-sitting is shown in Fig.7, the crest and trough of the depth curve corresponds to the time of floodtide and ebb.
Fig.6 Voyage near surface of landing AUV Fig.7 Data of bottom-sitting exploration Conclusion The landing AUV II is developed for long term survey of the temperature, salinity and current profile based on the improvement of the landing AUV I.
B., “Experimental verification of mission planning by autonomous mission execution and data visualization using the NPS AUV II,” in AUV ’92 IEEE.

In spite of the large amount of experimental data published in the literature, the influence of metallurgical parameters (original interlamellar spacing, pearlite colony size, alloying elements) and of processing parameters is not yet well established.
Mechanical properties were examined by tensile test measurements to establish some relationships between microstructural data and the mechanical properties of cold drawn pearlitic steels.
Our pictures are consistent with published data obtained on similar steel and show elongated lamellae that are parallel to the wire axis with an interlamellar spacing much lower than 50 nm.
These data indicate that the dissolution rate of Fe3C seems to be higher in steel 2 than in steel 1 even if few stoichiometric Fe3C clusters were also found within lamellae.
More generally, in collected 3D-AP data, the C concentration in former Fe3C lamellae was found to be in a range of 4 to 13 at.% in steel 3 after drawing and few stoichiometric Fe3C clusters were also found within lamellae.

Work Expert system works can be described as follows: For a large forgings, the first in the forging process the metal structure of dynamic test data, such as encoding parameters, encoding and fault category corresponding to the formation of learning samples, respectively, assigned to the knowledge base, through the learning samples to get the network weight matrix.
Accused of human-machine interface to the data object into a dynamic database, inference engine based on the knowledge in knowledge and dynamic real-time data in the database of reasoning, whether the system is obtained and what the failure fault, and then to evaluate the diagnostic results and decision-making, troubleshooting, and then repeat the process.
(3) Knowledge acquisition databases: reasoning used to store the initial data, intermediate results and final conclusions of the memory
Condition Conditions Conditional Conclusion 1 1 Module W / Module high <0.5 and irrational processes and Irrational process 2 W / module high <0.4 and 3 reduction rate <10% and Expert knowledge acquisition is to establish a sound system, sound and effective knowledge base.
Commonly used method has specific sort, sort order, data order, the nearest sorting, size sorting, the system uses the order of ranking method, combined with recombinant technology knowledge base, the rules of the higher frequency of use as a priority rule.

The beneficial effect of fluoride upon combination with sulfate compound seems to be attributed to reduction of the temperature of melt formation.
XRD data indicates that C2S and C3S peaks show low intensities at 1350℃ without addition of mineralizers (Fig. 1a).
After elimination of calcium silicate through the pretreatment process, XRD measurement was carried out and the data is shown in Fig. 2.
XRD data indicates that calcium langbeinite peaks show low intensities at 1350℃ without addition of SO3 (Fig. 2a).
: Langbeinite : 1/2-CaSO4 : CaO : K2SO4 Ref (1350℃) K20-0.25 (1350℃) K20-0.5 (1350℃) 10 20 30 40 50 60 2θ 10 20 30 40 50 60 2θ : C3A : C4AF : CaSO4 : MgO : Langbeinite : 1/2-CaSO4 : CaO Ref (1350℃) F-0.2+SO3-3.0 K20-0.25 (1350℃) F-0.2+SO3-3.0 K20-0.5 (1350℃) 10 20 30 40 50 60 2θ : C3A : C4AF : CaSO4 : MgO : Langbeinite : 1/2-CaSO4 : CaO Ref (1350℃) F-0.2+SO3-3.0 K20-0.25 (1350℃) F-0.2+SO3-3.0 K20-0.5 (1350℃) (a) No addition of mineralizer (b) Pair-mineralizer [F(0.2%)+SO3 (3.0%)] Fig.2 X-ray powder diffraction data with variation of K2O and pair-mineralizer contents.

The intelligent vehicle suppresses the interference from both the software and hardware for sampled path information to guarantee the reliability of data.
The intelligent vehicle uses reasonable algorithm for data processing and real-time returns bodywork offset guide line distance.
The speed detection system is composed of integrated hall sensors and 8 magnetic steels symmetrically placed on reduction gear driven by motor.
Due to the sensors are adjacent each other, the infrared radiation launched from the launch terminal of a sensor may be received by adjacent sensors by diffusing and it is possible to data error.
Then read again and compared data of these three times, the consistent data will be the path information in this cycle.

The milling experiments were carried out on rotor material by uniform design method, which can make the experiment data highly dispersed and representative.
In order to make the experiment data more evenly dispersed in the range, uniform design table U13* with a deviation of 0.1442 is chosen to arrange the experiments.
Feed of tooth mm/z Cutting Width mm Dry cutting Wet cutting Fx N Fy N Fz N Ra μm Fx N Fy N Fz N Ra μm 1 50 0.14 1.70 394.32 240.56 73.02 1.904 408.55 255.24 80.18 1.609 2 59 0.09 1.22 247.12 191.65 75.47 1.887 230.83 165.38 63.43 1.55 3 68 0.18 0.74 245.60 137.15 103.35 2.097 236.83 128.04 93.56 2.380 4 77 0.13 0.26 73.50 31.07 57.50 3.24 75.66 29.60 59.74 1.039 5 86 0.08 2.02 298.76 243.04 49.29 0.810 310.90 238.97 54.83 1.418 6 95 0.17 1.54 412.38 281.45 94.00 1.373 414.84 258.29 87.50 1.827 7 104 0.12 1.06 231.87 170.36 63.41 0.918 251.83 188.27 75.84 0.685 8 113 0.07 0.58 101.94 77.70 55.08 0.970 101.26 79.34 58.12 0.552 9 122 0.16 0.10 31.44 9.39 31.38 1.641 39.06 11.48 38.88 1.007 10 131 0.11 1.86 327.20 257.90 58.40 0.952 340.62 257.06 61.55 2.603 11 140 0.06 1.38 169.98 151.32 45.21 0.730 180.91 157.29 45.43 0.869 12 149 0.15 0.90 248.25 186.98 87.28 2.328 244.31 170.34 81.77 1.992 13 158 0.10 0.42 88.29 55.44 65.92 1.315 83.19 49.00 57.44 1.082 According to the data
Constant 3.150 40.079 0.000 -0.080 -0.040 -1.963 0.081 0.691 0.330 16.004 0.000 0.851 1.021 49.60 0.000 Similarly, the data about cutting force under the condition of dry cutting and the data about surface roughness were analyzed by SPSS statistics software.
The cutting fluid does not bring reduction to the cutting force.

In order to realize the carbon reduction, it has a certain theoretical and practical significance for the research of the relationship between land use change and soil carbon pools.
Data Collection and Analysis The data used to estimate land areas were extracted from remote sensing images.
The data of soil organic carbon from different types obtained from field laboratory experimental analysis soil samples.
Land use types were determined by a combination of supervised classification and manual interpretation of satellite images, supplemented with secondary information on geomorphology, historical vegetation maps, and ground truth data (Kilic et al., 2006).
This paper used both ERDAS Imagine 9.0 ,GPS and ARCGIS9.0 integrates the data using standard GIS features.

Data and Methods Data sources and processing.
In this study, landsat TM/ETM images covering the whole region, and whose imaging time were 2001.09.14(121-36, Landsat-7 ETM+), 2010.10.28(121-36, Landsat-5 TM) were selected as remote sensing data.
After radiometric correcting, geometric correcting and clipping, we carried out man-computer interactive interpretation and repeated modification until the data met the accuracy.
We evaluated the classification precision of images (taking 2010 as an example) using the land survey data in the same period and outdoor investigations of typical area.
As Table 2 shows, main reasons of farmland reduction were urbanization, increasing in industrial and mining land and surface subsidence leading to stagnant water.

This method was based on a fixed number of gyrations and was established on data from the virgin mixture and the 17% crumb rubber mixture at differing binder contents.
This is illustrated below in Fig. 1, for which a non-linear trend-line is shown to characterize the relationship most accurately for the available sets of data.
From an analysis of the data, a trend in variance was observed for the resilient modulus values between the load pulses in individual specimens.
The data from the test results suggest that with an increase in crumb rubber proportions, the variance in resilient modulus between loading pulses decreases as shown below in Fig. 2 and is quantified by standard deviation.
The remaining three data points in the mid-range correspond to the 20°C testing.