Development and evaluation of neural network freeway incident detection models using field data

This paper discusses a multi-layer feedforward (MLF) neural network incident detection model that was developed and evaluated using field data. In contrast to published neural network incident detection models which relied on simulated or limited field data for model development and testing, the model described in this paper was trained and tested on a real-world data set of 100 incidents. The model uses speed, flow and occupancy data measured at dual stations, averaged across all lanes and only from time interval t. The off-line performance of the model is reported under both incident and non-incident conditions. The incident detection performance of the model is reported based on a validation-test data set of 40 incidents that were independent of the 60 incidents used for training. The false alarm rates of the model are evaluated based on non-incident data that were collected from a freeway section which was video-taped for a period of 33 days. A comparative evaluation between the neural network model and the incident detection model in operation on Melbourne's freeways is also presented. The results of the comparative performance evaluation clearly demonstrate the substantial improvement in incident detection performance obtained by the neural network model. The paper also presents additional results that demonstrate how improvements in model performance can be achieved using variable decision thresholds. Finally, the model's fault-tolerance under conditions of corrupt or missing data is investigated and the impact of loop detector failure/malfunction on the performance of the trained model is evaluated and discussed. The results presented in this paper provide a comprehensive evaluation of the developed model and confirm that neural network models can provide fast and reliable incident detection on freeways. (C) 1997 Elsevier Science Ltd. All rights reserved.

Two-body and three-body abrasion: A critical discussion

It is argued that the common classification of abrasive wear into 'two-body abrasion' and 'three-body abrasion' is seriously flawed. No definitions have been agreed upon for these terms, and indeed there are two quite different interpretations, the implications of which are mutually inconsistent. In the dominant interpretation, the primary thrust of the two-body/three-body concept is to describe whether the abrasive particles are constrained (two-body) or free to roll (three-body). In this view, two-body abrasion is generally much more severe than three-body. The alternative interpretation emphasises the presence (three-body) or absence (two-body) of a rigid counterface backing the abrasive. In this view, three-body abrasion is equated to high-stress (or grinding) abrasion and is generally more severe than two-body (low-stress) abrasion. This paper recommends that the 'two-body/three-body' terminology be abandoned, to be replaced by an alternative classification scheme based directly upon the manifest severity of wear. (C) 1998 Elsevier Science S.A.

The critical power function is dependent on the duration of the predictive exercise tests chosen

The linear relationship between work accomplished (W-lim) and time to exhaustion (t(lim)) can be described by the equation: W-lim = a + CP.t(lim). Critical power (CP) is the slope of this line and is thought to represent a maximum rate of ATP synthesis without exhaustion, presumably an inherent characteristic of the aerobic energy system. The present investigation determined whether the choice of predictive tests would elicit significant differences in the estimated CP. Ten female physical education students completed, in random order and on consecutive days, five art-out predictive tests at preselected constant-power outputs. Predictive tests were performed on an electrically-braked cycle ergometer and power loadings were individually chosen so as to induce fatigue within approximately 1-10 mins. CP was derived by fitting the linear W-lim-t(lim) regression and calculated three ways: 1) using the first, third and fifth W-lim-t(lim) coordinates (I-135), 2) using coordinates from the three highest power outputs (I-123; mean t(lim) = 68-193 s) and 3) using coordinates from the lowest power outputs (I-345; mean t(lim) = 193-485 s). Repeated measures ANOVA revealed that CPI123 (201.0 +/- 37.9W) > CPI135 (176.1 +/- 27.6W) > CPI345 (164.0 +/- 22.8W) (P < 0.05). When the three sets of data were used to fit the hyperbolic Power-t(lim) regression, statistically significant differences between each CP were also found (P < 0.05). The shorter the predictive trials, the greater the slope of the W-lim-t(lim) regression; possibly because of the greater influence of 'aerobic inertia' on these trials. This may explain why CP has failed to represent a maximal, sustainable work rate. The present findings suggest that if CP is to represent the highest power output that an individual can maintain for a very long time without fatigue then CP should be calculated over a range of predictive tests in which the influence of aerobic inertia is minimised.

CO2 reforming of methane on Ni catalysts: Effects of the support phase and preparation technique

The effects of the support phase and catalyst preparation methods on catalytic activity and carbon deposition were systematically investigated over nickel catalysts supported on Al2O3, SiO2 and MgO for the reforming reaction of methane with carbon dioxide. It is found that the pore structure of the support and metal-support interaction significantly affected the catalytic activity and coking resistance. Catalyst with well-developed porosity exhibited higher catalytic activity. Strong interaction between metal and the support made the catalyst more resistant to sintering and coking, thus resulting in a longer time of catalyst stability. (C) 1998 Elsevier Science B.V.

Radiation use efficiency increases when the diffuse component of incident radiation is enhanced under shade

Theoretical analyses have shown the radiation use efficiency of maize, soybean, and peanut to increase with a decrease in the level of incident radiation and an increase in the proportion of diffuse radiation. This study compared the growth and radiation use efficiency of Panicum maximum cv. Petrie (green panic) and Bothriochloa insculpta cv. Bisset (creeping bluegrass) beneath shading treatments (birdguard and solarweave shadecloths) with that in full sunlight. A level of incident radiation reduced by 25% under birdguard shadecloth decreased final yield and final leaf area index, but increased canopy leaf nitrogen concentration and radiation use efficiency (19-14%) (compared with the full sun treatment). A similar level of reduced incident radiation under solarweave shadecloth (which provided an increased proportion of diffuse radiation), increased final yield and radiation use efficiency (46-50%). An understanding of the effects of composition of incident radiation on radiation use efficiency of tropical grasses enables more accurate estimation of potential pasture growth in shaded environments. It also has impact upon crop production in glasshouses and greenhouses.

Critical sets for families of Latin squares

In this paper I give details of new constructions for critical sets in latin squares. These latin squares, of order n, are such that they can be partitioned into four subsquares each of which is based on the addition table of the integers module n/2, an isotopism of this or a conjugate.

Effect of load distance on self-selected manual lifting technique

The aim of this research is to determine the effects of constraining the horizontal distance of the feet from the load on the posture adopted at the start of the lift. Kinematic data were collected while each of 24 subjects lifted 3, 6, and 9 kg loads from a starting height 18 cm above the ground. The position of the feet was controlled relative to the load such that the horizontal distance from the hand to the ankle at the start of extension was either 20, 40, or 60 cm. Subjects performed 20 trials in each of six combinations of load and ankle-load distance chosen to provide three sets of equivilent load moment pairs. The initial horizontal distance from the load to the ankle had a large influence on the posture adopted to lift the load. Ankle and knee flexion, in particular, were reduced when the ankle-load distance was smaller, and particularly so when the distance was reduced to 20 cm. Hip flexion was reduced to a smaller extent, while lumbar vertebral flexion remained relatively unchanged. The inclination of the trunk at the start of the lift was unchanged when the ankle-load distance was 60 or 40 cm, but was 10 degrees greater when the load was 20 cm from the ankles, indicating that subjects adopted a posture closer to a stoop when the ankle-load distance was small. Comparison of conditions of equal load moment (but different load mass and ankle-load distance) revealed differences which mirrored the effects of ankle-load distance alone, suggesting that the effects of ankle-load distance on the posture adopted at the start of extension were largely independent of the load moment. While the forces and torques required to lift a load must be to some extent dependent on the load moment, rather than load or ankle-load distance per se, the posture adopted to lift the load is not.

Critical values for unit root and cointegration test statistics - the use of response surface equations

This note considers the value of surface response equations which can be used to calculate critical values for a range of unit root and cointegration tests popular in applied economic research.

Effect of disorder on quantum phase transitions in anisotropic XY spin chains in a transverse field

We present some exact results for the effect of disorder on the critical properties of an anisotropic XY spin chain in a transverse held. The continuum limit of the corresponding fermion model is taken and in various cases results in a Dirac equation with a random mass. Exact analytic techniques can then be used to evaluate the density of states and the localization length. In the presence of disorder the ferromagnetic-paramagnetic or Ising transition of the model is in the same universality class as the random transverse field Ising model solved by Fisher using a real-space renormalization-group decimation technique (RSRGDT). If there is only randomness in the anisotropy of the magnetic exchange then the anisotropy transition (from a ferromagnet in the x direction to a ferromagnet in the y direction) is also in this universality class. However, if there is randomness in the isotropic part of the exchange or in the transverse held then in a nonzero transverse field the anisotropy transition is destroyed by the disorder. We show that in the Griffiths' phase near the Ising transition that the ground-state energy has an essential singularity. The results obtained for the dynamical critical exponent, typical correlation length, and for the temperature dependence of the specific heat near the Ising transition agree with the results of the RSRODT and numerical work. [S0163-1829(99)07125-8].

The crack tip strain field of AISI 4340 - Part I - Measurement technique

This is the first paper in a study on the influence of the environment on the crack tip strain field for AISI 4340. A stressing stage for the environmental scanning electron microscope (ESEM) was constructed which was capable of applying loads up to 60 kN to fracture-mechanics samples. The measurement of the crack tip strain field required preparation (by electron lithography or chemical etching) of a system of reference points spaced at similar to 5 mu m intervals on the sample surface, loading the sample inside an electron microscope, image processing procedures to measure the displacement at each reference point and calculation of the strain field. Two algorithms to calculate strain were evaluated. Possible sources of errors were calculation errors due to the algorithm, errors inherent in the image processing procedure and errors due to the limited precision of the displacement measurements. Estimation of the contribution of each source of error was performed. The technique allows measurement of the crack tip strain field over an area of 50 x 40 mu m with a strain precision better than +/- 0.02 at distances larger than 5 mu m from the crack tip. (C) 1999 Kluwer Academic Publishers.