I am high school student, I have doubt if I have to locate a planet or any star in space, how do I locate it, because we have no perfect coordinate system, I have read in some books, that time is define as the one of the coordinate, in coordinate system, can we define something only by time as coordinate.
For locating objects in the sky, the horizontal and equatorial coordinate systems are commonly used. These systems describe the position of some object in the sky very well, but do not explain the position of the object in space (if you know the distance you know "where" the object is, but this is relative to the equatorial/horizontal plane and is tricky to wrap your head around).
If you want to precisely locate the position of some object in space relative to the ecliptic, the ecliptic coordinate system exists for that, and for locating deep sky objects, the galactic coordinate system is normally used, as it nicely describes positions of objects relative to our galactic disc.
The space with the fourth coordinate is used for relativistic models (it fits well into special relativity), but not for describing positions of objects (to see more about that, look up Minkowski space).
...because we have no perfect coordinate system...
That's okay, there really is no such thing as a perfect anything (except in Mathematics).
You are right that it's necessary to consider time since everything is moving. One time system (used by SPICE, see below) is TDB or Barycentric Dynamical Time.
Here is a recent set of slides An Overview of Reference Frames and Coordinate Systems in the SPICE Context from Navigation and Ancillary Information Facility (NAIF) at NASA's Jet Propulsion Laboratory (JPL).
SPICE is one convention/environment for handling events, locations, times, and motion in space. You can read a little about the J2000.0 coordinate system . There are several different time scales to consider since there are relativistic effects related to relative motion and gravitational potential.
There is also a short tutorial on the Las Cumbres Observatory's page Cosmic Coordinates.
You can search for all of these terms in other questions and answers in this site, and and in Space Exploration SE as well.
The more specific your question is, the easier it will be to write a specific answer.
Motion of Planets
Yes, you can define a planets position with only time as a coordinate.
I think part of what you are asking for are orbital state vectors, the vectors start with simple linear trajectories easily defined by six parameters three for position and three for velocity at a single instant in time (not what you are asking about), the next level of detail involves many-body simulations using these 6 parameters (as well as the parameters of the other body's in the system and their mass) and are hard to extrapolate with time and would require dynamically simulating the effects of general relativity and this is where the Keplerian elements are useful by defining a Kepler orbit with respect to the orbits shape (Eccentricity which determines if it is, in order, a circle, ellipse, parabola, or hyperbola), width (semimajor-axis, or radius for circular orbits), tilt of the orbit with respect to a fixed celestial equator (Inclination), the rotation of the orbital plane with respect to a fixed celestial meridian (Longitude of the ascending node), the angle the ellipses(or hyperbolas) shape is rotated by along its inclined orbital plane (argument of periapsis), and last but not least the parameter you are interested in the true anomaly which gives the actual position of the mass in its Kepler orbit (these are only one of the ways to describe the shape and position of an orbit). However, these Keplerian elements are only enough to describe the shape and instantaneous position of an orbit, it still lacks velocity information, but they serve as an easy starting point for creating a formula that can properly describe an orbit without requiring full dynamic multi-body general relativity simulations.
The end results are the VSOP formulas which start with the basic elliptical Kepler orbits and modify them with time according to formulas to account for the effects of general relativity and the other planets and remain very accurate for thousands of years, and best of all they only require a single time coordinate to derive a position in the past or future.
Yes, you can define a stars position with only time as a coordinate, for most stars over a thousand years. You need only use the currently known position and velocity and extrapolate linearly with time and this will remain valid for about a thousand years (usually), simply because stars are much further apart and take much longer to collide and interact even for the fastest of stars. However this data will obviously be useless for binary stars or stars in very dense clusters. If you are trying to predict a stars position over many thousands of years the answer is no, and there is no replacement for many-body gravity simulations (unless you are willing to do the math and derive VSOP style formulas your self).