The Effect of Highway Vibration to The Hills Slope Stability by an Integrated GPS-Vibration Data Processing

. Bandung-Jakarta, Indonesia highway have high density vehicles pass even often jammed. High activity of vehicles cause repeated vibrations and occur continuously. In case, the highway that divides the hills. The passing vehicles give rise repeatedly vibration and indicated disturbance of hillside slope stability. Vehicles vibration depend on speed, weight, number of vehicles and road construction materials. A method in question is counting a vehicle volume and determine the volume pattern of the vehicle. The magnitude of vibration is function of the object distance from the sources. The magnitude of the vibration effect is very dependent on the geological structure of hills formation. Hilly slope stability indicates experiencing interference if the object or material slope changes position or displacement. GPS observation give position changes and its displacement. The effect of Toll road vibration to slope stability is known by GPS and Vibration simultaneous measurement. Data acquisition used six unit GPS receivers and six unit vibration sensor which is installed on the hillside beside the highway. Integrated data processing techniques using Empirical Mode Decomposition (EMD). Integrated GPS-Vibration data are implemented to find actual displacement which are caused by highway vibration.


Introduction
Cipularang toll road is toll roads on Java, Indonesia. Opened on 2005, this road connects the Jakarta-Cikampek Toll Road and the Padaleunyi Toll Road. The toll road is operated by PT Jasa Marga Tbk. The names come from abbreviation of Cikampek -Purwakarta -Padalarang. This toll road stretches along the mountains so the track is going up and downhill and many bends and bridges. Using the toll road can save up to half the time because a shortcut between Bandung and Jakarta.
The volume of various types and weights of vehicles crossing the road, even on weekends and holidays there is a traffic jam. December 2015 data the average vehicle that passes is 40 thousand per day and each year has increased [2]. Mid-2018, the number of vehicles examined is 259 units, 39% of the average daily traffic volume, see below Figure 1. Of that number, the vehicles that were proven to be overloaded 153 units (59.07%), over dimensions of 33 units (12.74%), and normal 73 units (28.19%). This road is the most populated toll road in Indonesia. In such conditions, the vehicle creates vibrations that take place continuously and repeatedly. This situation can trigger a disturbance in the surrounding slope stability. The volume of vehicles of the Cipularang toll road has increased steadily, see Figure 2. So monitoring the impact of vehicle vibrations on the slopes needs to be done.   [2] Therefore, to determine the effect of vehicle loads passing through the toll road on slope stability, GPS observations and vibration measurements were carried out simultaneously.

Data Acquisition
The monitoring points are installed on the slopes around the toll road, see Figure 3. GNSS observations are carried out simultaneously with vibration measurements. Data acquisition is done simultaneously so that time synchronization can be done well. First Period data acquisition was carried out at 6 monitoring points. Furthermore, the monitoring point is added to get a better distribution. The research area is between the Cipularang toll road (north) and the Darangdan Purwakarta highway (south). So the research area gets the vibrational influence of both.

GNSS Observation
GNSS observation used specification as follow:

Vibration Measurement
The instrument used is a triaxial vibration sensor. Vibration sensor measurement where placed at the ground of monitoring point use specification as follow: All real processes that must be faced in practice are complex, as a rule, which consists of a large number of components. A chart of this type is therefore quite difficult to be analyzed as its components, when interacting with each other, mask and distort the regularities we would like to identify. This gives rise to a rightful desire to break down the process under consideration into individual components and analyze each of the components separately [7]. Analysis of individual components and consideration of contributions made in the existing process will help to better understand the processes that are running, as well as, can improve the reliability of estimates.
The term "decomposition" formally means the breaking down of a compound process or a composite material into separate constituent components. But in many areas related to analysis of different processes, signal analysis, analysis of various sorts of sequences. These functions are sort of artificially formed in the process of data decomposition but despite their "artificial" origin they allow for a deeper analysis of data helping to identify hidden patterns. There is a lot of various decomposition methods that can be applied in practice to a given sequence under consideration. These methods may have different underlying mathematical or empirical approaches, different degree of complexity and different areas of application [7].
The algorithm as proposed by Huang is based on producing smooth envelopes defined by local maxima and minima of a sequence and subsequent subtraction of the mean of these envelopes from the initial sequence. This requires the identification of all local extrema that are further connected by cubic spline lines to produce the upper and the lower envelopes. The procedure of plotting the envelopes is shown in Figure 4. Emd decomposes a signal X(t) into k number of intrinsic mode functions (IMF), and residual rk(t) using the sifting process. A brief overview of the sifting process, listed in

GNSS Data Processing
GNSS data processing uses a kinematic method. GNSS survey data for 30-60 minutes is processed as initiation. after that data processing is done by kinematic algorithm. the goal is that we can get positions every 5 seconds. The results of GNSS processing use a topocentric coordinate system (east, north, up) and find accelerations to match the vibration measurement system. In order for the analysis to be carried out, the suitability of GNSS observation and vibration time must be the same, following the intended time suitability.

Vibration Data Processing
Vibration measurement uses a triaxial vibration sensor so that the acceleration data has a value of 3 axes (x, y, z). The unit is m/s2. For analysis, the data process is carried out for each axis, as well as gnss data. The following are the results of processing for vibrations data at kbn1 and msjd97 point.
Plotting of kbn1 vibrations data (x, y, z)

Discussion and Conclusion
The following conclusions are based on the assumption that the observations of GNSS and Vibration are carried out for 2 periods and at the same time. For further studies, indications of the conclusions above will be examined further. .